{
  "analytic_examples": {
    "two_step_first_logit_gradient": {
      "on_policy": 0.1,
      "unweighted": -0.1,
      "action_only_is": -0.1,
      "trajectory_is": 0.1
    },
    "constant_reward_logit_gradient": {
      "behavior_probability": 0.1,
      "target_probability": 0.9,
      "reward_minus_baseline": 1.0,
      "on_policy": 0.0,
      "unweighted": -0.8,
      "action_is": 0.0,
      "interpretation": "Action-independent reward with a nonzero baseline error; an exact baseline removes this term."
    },
    "independent_bernoulli_trajectory_weights": [
      {
        "behavior_probability": 0.5,
        "target_probability": 0.55,
        "horizon": 1,
        "population_ess_fraction": 0.9900990099009901
      },
      {
        "behavior_probability": 0.5,
        "target_probability": 0.55,
        "horizon": 20,
        "population_ess_fraction": 0.8195444703372954
      },
      {
        "behavior_probability": 0.5,
        "target_probability": 0.55,
        "horizon": 256,
        "population_ess_fraction": 0.07829396937632106
      },
      {
        "behavior_probability": 0.6,
        "target_probability": 0.8,
        "horizon": 1,
        "population_ess_fraction": 0.8571428571428571
      },
      {
        "behavior_probability": 0.6,
        "target_probability": 0.8,
        "horizon": 20,
        "population_ess_fraction": 0.04582096247806903
      },
      {
        "behavior_probability": 0.6,
        "target_probability": 0.8,
        "horizon": 256,
        "population_ess_fraction": 7.271463936932875e-18
      },
      {
        "behavior_probability": 0.1,
        "target_probability": 0.9,
        "horizon": 1,
        "population_ess_fraction": 0.1232876712328767
      },
      {
        "behavior_probability": 0.1,
        "target_probability": 0.9,
        "horizon": 20,
        "population_ess_fraction": 6.5825321111925685e-19
      },
      {
        "behavior_probability": 0.1,
        "target_probability": 0.9,
        "horizon": 256,
        "population_ess_fraction": 1.8855160154680256e-233
      }
    ]
  },
  "profiles": {
    "territory/ippo": {
      "checkpoints": [
        {
          "checkpoint": "gridworlds/territory/role_claim/2/ippo/checkpoints/run_1_seed_0.pkl",
          "checkpoint_sha256": "538acd8c50cdd7a461666cfdbab61521d1ca873c79113f9552c778013b7093b8",
          "run_seed": 0,
          "config": {
            "NUM_ENVS": 1,
            "TOTAL_TIMESTEPS": 250112,
            "NUM_STEPS": 256,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 1.0,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 0,
            "AUTO_RESUME_EXTRA_TIMESTEPS": 49920
          },
          "episodes": 8,
          "diagnostic_seed0": 812000000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed & ~player_1_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.4327841591787731e-05,
                  "mean_kl_behavior_to_synthetic": 2.298297769844594e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999935116448,
                    "ess_fraction": 0.9999999968318578,
                    "log_weight_quantiles": [
                      -0.0003075599670410156,
                      -6.325244903564452e-05,
                      0.0,
                      5.8007240295410145e-05,
                      0.00027251243591308594
                    ],
                    "log_mean_weight": 8.848809222226873e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.99998978153682,
                    "ess_fraction": 0.9999987226921025,
                    "log_weight_quantiles": [
                      -0.0021524429321289062,
                      -0.0008409976959228515,
                      1.5497207641601562e-06,
                      0.001011240482330322,
                      0.0022890567779541016
                    ],
                    "log_mean_weight": 2.2886029038046447e-05
                  }
                },
                "2": {
                  "mean_tv": 1.9397152434400393e-05,
                  "mean_kl_behavior_to_synthetic": 3.3105442704751187e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.999987936732,
                    "ess_fraction": 0.9999999941097324,
                    "log_weight_quantiles": [
                      -0.00045680999755859375,
                      -6.45637512207031e-05,
                      0.0,
                      7.996559143066403e-05,
                      0.0004646778106689453
                    ],
                    "log_mean_weight": -8.486328878535773e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999996621714659,
                    "ess_fraction": 0.9999995777143323,
                    "log_weight_quantiles": [
                      -0.0012564659118652344,
                      -0.0010007858276367188,
                      -0.00028973817825317383,
                      0.0005502581596374511,
                      0.0007007122039794922
                    ],
                    "log_mean_weight": -0.00021779283923206022
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed & ~player_0_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 2.8708370967690763e-06,
                  "mean_kl_behavior_to_synthetic": 4.720221923164798e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999997711589,
                    "ess_fraction": 0.9999999998882612,
                    "log_weight_quantiles": [
                      -5.4836273193359375e-05,
                      -1.3470649719238281e-05,
                      0.0,
                      1.0967254638671875e-05,
                      4.9591064453125e-05
                    ],
                    "log_mean_weight": 1.745996041863996e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999736637655,
                    "ess_fraction": 0.9999999670797068,
                    "log_weight_quantiles": [
                      -0.00043904781341552734,
                      -0.00015316009521484372,
                      4.506111145019531e-05,
                      0.00014262199401855467,
                      0.00018334388732910156
                    ],
                    "log_mean_weight": 4.4719105898342534e-06
                  }
                },
                "2": {
                  "mean_tv": 5.29007295751134e-06,
                  "mean_kl_behavior_to_synthetic": 4.743548036394849e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.999999395361,
                    "ess_fraction": 0.9999999997047662,
                    "log_weight_quantiles": [
                      -6.771087646484375e-05,
                      -1.621246337890625e-05,
                      0.0,
                      2.129077911376952e-05,
                      9.775161743164062e-05
                    ],
                    "log_mean_weight": 2.127699538031927e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999183276708,
                    "ess_fraction": 0.9999998979095885,
                    "log_weight_quantiles": [
                      -0.00037384033203125,
                      -0.0003378748893737793,
                      -9.828805923461914e-05,
                      0.0004248619079589843,
                      0.0005784034729003906
                    ],
                    "log_mean_weight": 5.460161116545227e-06
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "gridworlds/territory/role_claim/2/ippo/checkpoints/run_2_seed_1.pkl",
          "checkpoint_sha256": "1d3c20c4d739c6cdfdba6638209f7b6c78b7c2387676faf8ae4f3dcc8135e2df",
          "run_seed": 1,
          "config": {
            "NUM_ENVS": 1,
            "TOTAL_TIMESTEPS": 200192,
            "NUM_STEPS": 256,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 1.0,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 1
          },
          "episodes": 8,
          "diagnostic_seed0": 812010000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed & ~player_1_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 5.318200833102177e-06,
                  "mean_kl_behavior_to_synthetic": -2.0625551563738134e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.999999498002,
                    "ess_fraction": 0.9999999997548837,
                    "log_weight_quantiles": [
                      -8.7738037109375e-05,
                      -1.5914440155029297e-05,
                      -1.6093254089355469e-06,
                      1.811981201171875e-05,
                      8.916854858398438e-05
                    ],
                    "log_mean_weight": -1.510718405658371e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.99999947379869,
                    "ess_fraction": 0.9999999342248362,
                    "log_weight_quantiles": [
                      -0.0006406307220458984,
                      -0.00026545524597167965,
                      -1.7911195755004883e-05,
                      0.0001985669136047363,
                      0.00025773048400878906
                    ],
                    "log_mean_weight": -3.8672871155028064e-05
                  }
                },
                "2": {
                  "mean_tv": 1.8521001310655605e-07,
                  "mean_kl_behavior_to_synthetic": -6.991175811326565e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999999746597,
                    "ess_fraction": 0.9999999999876268,
                    "log_weight_quantiles": [
                      -7.033348083496094e-05,
                      -1.1920928955078125e-07,
                      0.0,
                      1.1920928955078125e-07,
                      4.887580871582031e-05
                    ],
                    "log_mean_weight": 3.5018094594604714e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999958921155,
                    "ess_fraction": 0.9999999948651443,
                    "log_weight_quantiles": [
                      -0.00010842084884643555,
                      -8.305311203002929e-05,
                      7.838010787963867e-06,
                      9.192824363708496e-05,
                      0.00011658668518066406
                    ],
                    "log_mean_weight": 8.965615904328806e-06
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed & ~player_0_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.4386193888150595e-05,
                  "mean_kl_behavior_to_synthetic": 3.385990091703327e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999950829517,
                    "ess_fraction": 0.9999999975990975,
                    "log_weight_quantiles": [
                      -0.0002110004425048828,
                      -5.840063095092773e-05,
                      0.0,
                      6.492137908935546e-05,
                      0.00017786026000976562
                    ],
                    "log_mean_weight": 3.0251241197744414e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999996412890511,
                    "ess_fraction": 0.9999995516113139,
                    "log_weight_quantiles": [
                      -0.0008602142333984375,
                      -0.0007573246955871582,
                      7.843971252441406e-05,
                      0.001022970676422119,
                      0.0010464191436767578
                    ],
                    "log_mean_weight": 7.736004092292472e-05
                  }
                },
                "2": {
                  "mean_tv": 2.5559941989185345e-06,
                  "mean_kl_behavior_to_synthetic": 4.168833302227592e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999989693447,
                    "ess_fraction": 0.9999999994967503,
                    "log_weight_quantiles": [
                      -0.00020694732666015625,
                      -4.76837158203125e-07,
                      0.0,
                      2.384185791015625e-07,
                      0.0002372264862060547
                    ],
                    "log_mean_weight": 2.1026486291149653e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.9999998362502485,
                    "ess_fraction": 0.9999999795312811,
                    "log_weight_quantiles": [
                      -0.0002269148826599121,
                      -7.128715515136717e-05,
                      2.1159648895263672e-05,
                      0.00022453665733337402,
                      0.00025278329849243164
                    ],
                    "log_mean_weight": 5.377362435132535e-05
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "gridworlds/territory/role_claim/2/ippo/checkpoints/run_3_seed_2.pkl",
          "checkpoint_sha256": "5130b40d620378bd56f754c21726e7a55f346e511a8fece60afdb0d6375ade29",
          "run_seed": 2,
          "config": {
            "NUM_ENVS": 1,
            "TOTAL_TIMESTEPS": 200192,
            "NUM_STEPS": 256,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 1.0,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 2
          },
          "episodes": 8,
          "diagnostic_seed0": 812020000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed & ~player_1_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.8445671451416493e-06,
                  "mean_kl_behavior_to_synthetic": 4.7374720371321944e-11,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999999317347,
                    "ess_fraction": 0.9999999999666673,
                    "log_weight_quantiles": [
                      -2.9087066650390625e-05,
                      -6.9141387939453125e-06,
                      0.0,
                      6.2167644500732395e-06,
                      2.9087066650390625e-05
                    ],
                    "log_mean_weight": -1.8304642703092575e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999950624477,
                    "ess_fraction": 0.9999999938280596,
                    "log_weight_quantiles": [
                      -0.00016611814498901367,
                      -0.0001572728157043457,
                      -2.9087066650390625e-05,
                      2.4312734603881818e-05,
                      9.077787399291992e-05
                    ],
                    "log_mean_weight": -4.6861065977277526e-05
                  }
                },
                "2": {
                  "mean_tv": 2.305258422353437e-06,
                  "mean_kl_behavior_to_synthetic": -4.107004931206983e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999998802884,
                    "ess_fraction": 0.9999999999415471,
                    "log_weight_quantiles": [
                      -2.193450927734375e-05,
                      -7.927417755126953e-06,
                      0.0,
                      7.62939453125e-06,
                      5.698204040527344e-05
                    ],
                    "log_mean_weight": 1.4790578911678427e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999813527698,
                    "ess_fraction": 0.9999999766909623,
                    "log_weight_quantiles": [
                      -0.00023251771926879883,
                      -0.00015524625778198242,
                      5.3554773330688477e-05,
                      0.00020694732666015625,
                      0.00023323297500610352
                    ],
                    "log_mean_weight": 3.7868054984377874e-05
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed & ~player_0_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 9.70178762620939e-07,
                  "mean_kl_behavior_to_synthetic": 2.241599249213825e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999999734089,
                    "ess_fraction": 0.999999999987016,
                    "log_weight_quantiles": [
                      -1.8835067749023438e-05,
                      -4.208087921142576e-06,
                      0.0,
                      3.2365322113037082e-06,
                      2.193450927734375e-05
                    ],
                    "log_mean_weight": 9.264398431974992e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999985579221,
                    "ess_fraction": 0.9999999981974026,
                    "log_weight_quantiles": [
                      -6.443262100219727e-05,
                      -9.733438491821283e-06,
                      2.5272369384765625e-05,
                      5.9920549392700186e-05,
                      0.0001016855239868164
                    ],
                    "log_mean_weight": 2.3716099347512233e-05
                  }
                },
                "2": {
                  "mean_tv": 1.064940787811763e-06,
                  "mean_kl_behavior_to_synthetic": -1.797980483789216e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999999672646,
                    "ess_fraction": 0.9999999999840159,
                    "log_weight_quantiles": [
                      -1.9550323486328125e-05,
                      -2.86102294921875e-06,
                      0.0,
                      3.3974647521972656e-06,
                      3.170967102050781e-05
                    ],
                    "log_mean_weight": 2.849344921791811e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999798542776,
                    "ess_fraction": 0.999999974817847,
                    "log_weight_quantiles": [
                      -4.380941390991211e-05,
                      -3.579854965209961e-05,
                      -9.5367431640625e-07,
                      0.00022903084754943842,
                      0.00046581029891967773
                    ],
                    "log_mean_weight": 7.295377276663649e-05
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "gridworlds/territory/role_claim/2/ippo/checkpoints/run_4_seed_3.pkl",
          "checkpoint_sha256": "56875cfcf3f129053b41ef42eaf19a2a51495659f8ab935d0106fd3408a186b5",
          "run_seed": 3,
          "config": {
            "NUM_ENVS": 1,
            "TOTAL_TIMESTEPS": 200192,
            "NUM_STEPS": 256,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 1.0,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 3
          },
          "episodes": 8,
          "diagnostic_seed0": 812030000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed & ~player_1_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 3.4833281034155833e-06,
                  "mean_kl_behavior_to_synthetic": 6.537350969075033e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.999999804942,
                    "ess_fraction": 0.9999999999047569,
                    "log_weight_quantiles": [
                      -3.8623809814453125e-05,
                      -1.1682510375976562e-05,
                      0.0,
                      1.1920928955078125e-05,
                      3.62396240234375e-05
                    ],
                    "log_mean_weight": 2.0668481797034776e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999942492939,
                    "ess_fraction": 0.9999999928116173,
                    "log_weight_quantiles": [
                      -9.143352508544922e-05,
                      -4.887580871582031e-05,
                      5.042552947998047e-05,
                      0.00014624595642089843,
                      0.00016510486602783203
                    ],
                    "log_mean_weight": 5.290271647805972e-05
                  }
                },
                "2": {
                  "mean_tv": 2.730086089710448e-06,
                  "mean_kl_behavior_to_synthetic": 4.630819517259003e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999997891982,
                    "ess_fraction": 0.9999999998970694,
                    "log_weight_quantiles": [
                      -7.581710815429688e-05,
                      -1.049041748046875e-05,
                      0.0,
                      1.1801719665527344e-05,
                      6.341934204101562e-05
                    ],
                    "log_mean_weight": 2.7519907830536106e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999535283711,
                    "ess_fraction": 0.9999999419104639,
                    "log_weight_quantiles": [
                      -0.00016033649444580078,
                      -0.00011944770812988281,
                      1.7464160919189453e-05,
                      0.000303661823272705,
                      0.0006567239761352539
                    ],
                    "log_mean_weight": 7.046682621448652e-05
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed & ~player_0_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.62751507879484e-05,
                  "mean_kl_behavior_to_synthetic": 3.826443787848674e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.99999549408,
                    "ess_fraction": 0.9999999977998437,
                    "log_weight_quantiles": [
                      -0.00015664100646972656,
                      -5.84721565246582e-05,
                      0.0,
                      6.628036499023438e-05,
                      0.00022721290588378906
                    ],
                    "log_mean_weight": 1.0127531619862247e-09
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999995425651824,
                    "ess_fraction": 0.999999428206478,
                    "log_weight_quantiles": [
                      -0.0015611052513122559,
                      -0.0008537709712982176,
                      0.00015047192573547363,
                      0.0006280481815338133,
                      0.0012227296829223633
                    ],
                    "log_mean_weight": 2.6362988515024437e-07
                  }
                },
                "2": {
                  "mean_tv": 1.2208774632095955e-05,
                  "mean_kl_behavior_to_synthetic": 4.038401349440626e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999959159795,
                    "ess_fraction": 0.9999999980058494,
                    "log_weight_quantiles": [
                      -0.00027871131896972656,
                      -4.867315292358398e-05,
                      0.0,
                      6.091594696044922e-05,
                      0.00016158819198608398
                    ],
                    "log_mean_weight": 1.24361423663649e-06
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999998839134901,
                    "ess_fraction": 0.9999998548918626,
                    "log_weight_quantiles": [
                      -0.0001722574234008789,
                      -0.00014075636863708496,
                      0.0002739429473876953,
                      0.000818634033203125,
                      0.0010184049606323242
                    ],
                    "log_mean_weight": 0.00031818253485702886
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "gridworlds/territory/role_claim/2/ippo/checkpoints/run_5_seed_4.pkl",
          "checkpoint_sha256": "7156286c87afafe03789b2a8fd2a3cfbd3f7b59244ef630b09cd9994efc0461d",
          "run_seed": 4,
          "config": {
            "NUM_ENVS": 1,
            "TOTAL_TIMESTEPS": 200192,
            "NUM_STEPS": 256,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 1.0,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 4
          },
          "episodes": 8,
          "diagnostic_seed0": 812040000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed & ~player_1_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 5.4948786936109905e-06,
                  "mean_kl_behavior_to_synthetic": -8.061706243049773e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999993897125,
                    "ess_fraction": 0.999999999702008,
                    "log_weight_quantiles": [
                      -8.392333984375e-05,
                      -1.8596649169921875e-05,
                      0.0,
                      2.1457672119140625e-05,
                      6.115436553955078e-05
                    ],
                    "log_mean_weight": 9.53185218610864e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999773615293,
                    "ess_fraction": 0.9999999717019116,
                    "log_weight_quantiles": [
                      -0.0002415180206298828,
                      -0.00013228654861450193,
                      8.106231689453125e-06,
                      0.00021789073944091794,
                      0.0003694295883178711
                    ],
                    "log_mean_weight": 2.437754665716497e-05
                  }
                },
                "2": {
                  "mean_tv": 2.8413562209290126e-06,
                  "mean_kl_behavior_to_synthetic": -1.4826819306590063e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999997446191,
                    "ess_fraction": 0.9999999998753023,
                    "log_weight_quantiles": [
                      -4.220008850097656e-05,
                      -1.2874603271484375e-05,
                      0.0,
                      8.022785186767573e-06,
                      9.059906005859375e-05
                    ],
                    "log_mean_weight": -1.4889105610628667e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999713389881,
                    "ess_fraction": 0.9999999641737352,
                    "log_weight_quantiles": [
                      -0.00040078163146972656,
                      -0.0002504944801330566,
                      3.534555435180664e-05,
                      0.00012818574905395505,
                      0.00023174285888671875
                    ],
                    "log_mean_weight": -3.8114157124477554e-05
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed & ~player_0_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 3.754834517384719e-05,
                  "mean_kl_behavior_to_synthetic": 7.51257851950997e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999553989835,
                    "ess_fraction": 0.9999999782221599,
                    "log_weight_quantiles": [
                      -0.0014560222625732422,
                      -0.00010042786598205566,
                      5.960464477539063e-08,
                      0.00013260841369628905,
                      0.0004200935363769531
                    ],
                    "log_mean_weight": -9.483989886503513e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999837578210612,
                    "ess_fraction": 0.9999796972763265,
                    "log_weight_quantiles": [
                      -0.010931015014648438,
                      -0.00362064242362976,
                      0.0003160238265991211,
                      0.003183728456497192,
                      0.0062018632888793945
                    ],
                    "log_mean_weight": -0.00023538816467672061
                  }
                },
                "2": {
                  "mean_tv": 2.2154771231647533e-05,
                  "mean_kl_behavior_to_synthetic": 5.460650104326677e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999787119489,
                    "ess_fraction": 0.9999999896054438,
                    "log_weight_quantiles": [
                      -0.00031495094299316406,
                      -9.468793869018554e-05,
                      0.0,
                      8.24928283691404e-06,
                      0.0008311271667480469
                    ],
                    "log_mean_weight": -2.26528099365954e-06
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999980984505707,
                    "ess_fraction": 0.9999976230632134,
                    "log_weight_quantiles": [
                      -0.003775179386138916,
                      -0.0023997724056243896,
                      -0.00015103816986083984,
                      0.0008029103279113768,
                      0.001652061939239502
                    ],
                    "log_mean_weight": -0.0005800527750404361
                  }
                }
              }
            }
          }
        }
      ],
      "primary": {
        "seeds": [
          {
            "eval_seed0": 503,
            "p_hat": {
              "player_0": 0.375,
              "player_1": 0.601
            },
            "minimum_p_hat": 0.375,
            "sum_p_hat": 0.976,
            "jointly_certified": true
          },
          {
            "eval_seed0": 1000503,
            "p_hat": {
              "player_0": 0.102,
              "player_1": 0.871
            },
            "minimum_p_hat": 0.102,
            "sum_p_hat": 0.973,
            "jointly_certified": false
          },
          {
            "eval_seed0": 2000503,
            "p_hat": {
              "player_0": 0.659,
              "player_1": 0.328
            },
            "minimum_p_hat": 0.328,
            "sum_p_hat": 0.9870000000000001,
            "jointly_certified": true
          },
          {
            "eval_seed0": 3000503,
            "p_hat": {
              "player_0": 0.426,
              "player_1": 0.551
            },
            "minimum_p_hat": 0.426,
            "sum_p_hat": 0.9770000000000001,
            "jointly_certified": true
          },
          {
            "eval_seed0": 4000503,
            "p_hat": {
              "player_0": 0.451,
              "player_1": 0.531
            },
            "minimum_p_hat": 0.451,
            "sum_p_hat": 0.982,
            "jointly_certified": true
          }
        ],
        "jointly_certified_count": 4,
        "mean_minimum_p_hat": 0.3364,
        "mean_sum_p_hat": 0.9790000000000001
      },
      "confirmatory": {
        "seeds": [
          {
            "eval_seed0": 500503,
            "p_hat": {
              "player_0": 0.379,
              "player_1": 0.598
            },
            "minimum_p_hat": 0.379,
            "sum_p_hat": 0.977,
            "jointly_certified": true
          },
          {
            "eval_seed0": 1500503,
            "p_hat": {
              "player_0": 0.102,
              "player_1": 0.871
            },
            "minimum_p_hat": 0.102,
            "sum_p_hat": 0.973,
            "jointly_certified": false
          },
          {
            "eval_seed0": 2500503,
            "p_hat": {
              "player_0": 0.66,
              "player_1": 0.327
            },
            "minimum_p_hat": 0.327,
            "sum_p_hat": 0.9870000000000001,
            "jointly_certified": true
          },
          {
            "eval_seed0": 3500503,
            "p_hat": {
              "player_0": 0.427,
              "player_1": 0.55
            },
            "minimum_p_hat": 0.427,
            "sum_p_hat": 0.9770000000000001,
            "jointly_certified": true
          },
          {
            "eval_seed0": 4500503,
            "p_hat": {
              "player_0": 0.449,
              "player_1": 0.533
            },
            "minimum_p_hat": 0.449,
            "sum_p_hat": 0.982,
            "jointly_certified": true
          }
        ],
        "jointly_certified_count": 4,
        "mean_minimum_p_hat": 0.33680000000000004,
        "mean_sum_p_hat": 0.9792000000000002
      }
    },
    "territory/ippo_cer": {
      "checkpoints": [
        {
          "checkpoint": "gridworlds/territory/role_claim/2/ippo_cer/checkpoints/run_1_seed_0.pkl",
          "checkpoint_sha256": "ddcaedf52a0e8e591d6c24f127a69c5a7553acad65a9d7c3f22052d4ac98c752",
          "run_seed": 0,
          "config": {
            "NUM_ENVS": 1,
            "TOTAL_TIMESTEPS": 200192,
            "NUM_STEPS": 256,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 1.0,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 0,
            "COUNTERFACTUAL_REPLAY": true
          },
          "episodes": 8,
          "diagnostic_seed0": 812000000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed & ~player_1_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 4.5488829230156106e-06,
                  "mean_kl_behavior_to_synthetic": 1.9145280426020153e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999996659633,
                    "ess_fraction": 0.9999999998368961,
                    "log_weight_quantiles": [
                      -6.079673767089844e-05,
                      -1.621246337890625e-05,
                      0.0,
                      1.800060272216797e-05,
                      6.103515625e-05
                    ],
                    "log_mean_weight": -6.778858079260948e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999938429574,
                    "ess_fraction": 0.9999999923036967,
                    "log_weight_quantiles": [
                      -0.0001342296600341797,
                      -9.784698486328124e-05,
                      -4.589557647705078e-05,
                      7.647275924682614e-05,
                      0.00016367435455322266
                    ],
                    "log_mean_weight": -1.737090597575479e-05
                  }
                },
                "2": {
                  "mean_tv": 2.0645229565990193e-06,
                  "mean_kl_behavior_to_synthetic": -1.6466937947486309e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999997821917,
                    "ess_fraction": 0.9999999998936483,
                    "log_weight_quantiles": [
                      -0.00011801719665527344,
                      -5.245208740234375e-06,
                      0.0,
                      7.867813110351562e-06,
                      5.924701690673828e-05
                    ],
                    "log_mean_weight": -2.2480301796628691e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999606013736,
                    "ess_fraction": 0.999999950751717,
                    "log_weight_quantiles": [
                      -0.0005844831466674805,
                      -0.00027689933776855465,
                      -1.7881393432617188e-05,
                      0.0001353025436401367,
                      0.00021374225616455078
                    ],
                    "log_mean_weight": -5.7538556451921044e-05
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed",
                      2
                    ],
                    [
                      "player_1_resource_claimed & ~player_0_resource_claimed",
                      1
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 6.669691473279531e-06,
                  "mean_kl_behavior_to_synthetic": 3.2069840044487117e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999970757092,
                    "ess_fraction": 0.9999999985721236,
                    "log_weight_quantiles": [
                      -0.00018584728240966797,
                      -3.2937526702880854e-05,
                      0.0,
                      3.099441528320204e-07,
                      0.000293731689453125
                    ],
                    "log_mean_weight": 1.2740646896489843e-06
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999994196270142,
                    "ess_fraction": 0.9999992745337678,
                    "log_weight_quantiles": [
                      -0.00084686279296875,
                      -0.00030554533004760736,
                      0.00010800361633300781,
                      0.001384782791137695,
                      0.002192378044128418
                    ],
                    "log_mean_weight": 0.0003263403226743324
                  }
                },
                "2": {
                  "mean_tv": 1.6114138799089953e-05,
                  "mean_kl_behavior_to_synthetic": 2.7536048307107958e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.999994072483,
                    "ess_fraction": 0.9999999971057046,
                    "log_weight_quantiles": [
                      -0.0003616809844970703,
                      -6.864070892333984e-05,
                      1.2636184692382812e-05,
                      3.8504600524902344e-05,
                      0.00011789798736572266
                    ],
                    "log_mean_weight": -9.947186193081818e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999997482437132,
                    "ess_fraction": 0.9999996853046415,
                    "log_weight_quantiles": [
                      -0.0010483264923095703,
                      -0.0007395744323730469,
                      -0.00046640634536743164,
                      0.00048427581787109365,
                      0.0008454322814941406
                    ],
                    "log_mean_weight": -0.00025486116453019665
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "gridworlds/territory/role_claim/2/ippo_cer/checkpoints/run_2_seed_1.pkl",
          "checkpoint_sha256": "88ecea3936ed6fdfe4fc41bef6f1580aa8660a5274810bcd4ffe217cbcb31d7f",
          "run_seed": 1,
          "config": {
            "NUM_ENVS": 1,
            "TOTAL_TIMESTEPS": 200192,
            "NUM_STEPS": 256,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 1.0,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 1,
            "COUNTERFACTUAL_REPLAY": true
          },
          "episodes": 8,
          "diagnostic_seed0": 812010000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed & ~player_1_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 3.356332414447949e-06,
                  "mean_kl_behavior_to_synthetic": 7.516254314530421e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999996170177,
                    "ess_fraction": 0.9999999998129969,
                    "log_weight_quantiles": [
                      -0.00011491775512695312,
                      -9.608268737792965e-06,
                      1.4603137969970703e-06,
                      8.106231689453125e-06,
                      8.20159912109375e-05
                    ],
                    "log_mean_weight": -4.116965931548423e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999730804934,
                    "ess_fraction": 0.9999999663506167,
                    "log_weight_quantiles": [
                      -0.0005240440368652344,
                      -0.00032176971435546873,
                      -3.4481287002563477e-05,
                      3.4874677658081036e-05,
                      0.00010442733764648438
                    ],
                    "log_mean_weight": -0.00010540143754388857
                  }
                },
                "2": {
                  "mean_tv": 9.895173536352855e-07,
                  "mean_kl_behavior_to_synthetic": -2.2418018033832423e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999998990074,
                    "ess_fraction": 0.9999999999506872,
                    "log_weight_quantiles": [
                      -3.910064697265625e-05,
                      -4.76837158203125e-06,
                      0.0,
                      2.384185791015625e-07,
                      0.00015211105346679688
                    ],
                    "log_mean_weight": -7.587813454786691e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999935585432,
                    "ess_fraction": 0.999999991948179,
                    "log_weight_quantiles": [
                      -0.00021392107009887695,
                      -8.991956710815429e-05,
                      -1.3232231140136719e-05,
                      5.626082420349119e-05,
                      0.0001360774040222168
                    ],
                    "log_mean_weight": -1.9427088139286047e-05
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed",
                      2
                    ],
                    [
                      "player_1_resource_claimed & ~player_0_resource_claimed",
                      1
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 0.000263274296120121,
                  "mean_kl_behavior_to_synthetic": 5.055323531927752e-07,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9980585578814,
                    "ess_fraction": 0.9999990520302156,
                    "log_weight_quantiles": [
                      -0.003838062286376953,
                      -0.0009382724761962872,
                      0.0,
                      0.0012138783931732178,
                      0.0037543773651123047
                    ],
                    "log_mean_weight": 2.4409484680642385e-05
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.997246795532795,
                    "ess_fraction": 0.9996558494415994,
                    "log_weight_quantiles": [
                      -0.0067969560623168945,
                      -0.004248374700546264,
                      -0.0029972195625305176,
                      0.02246736884117126,
                      0.05241811275482178
                    ],
                    "log_mean_weight": 0.0062955447551718835
                  }
                },
                "2": {
                  "mean_tv": 0.000525361993520724,
                  "mean_kl_behavior_to_synthetic": 1.2607758072306605e-06,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9947558855517,
                    "ess_fraction": 0.9999974393972421,
                    "log_weight_quantiles": [
                      -0.006552934646606445,
                      -0.0015734255313873291,
                      0.0,
                      0.002919840812683105,
                      0.007511615753173828
                    ],
                    "log_mean_weight": 2.658509393890264e-05
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.996006135479312,
                    "ess_fraction": 0.999500766934914,
                    "log_weight_quantiles": [
                      -0.03069460391998291,
                      -0.023372966051101684,
                      0.007739067077636719,
                      0.02806041836738586,
                      0.044422149658203125
                    ],
                    "log_mean_weight": 0.006728465943314568
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "gridworlds/territory/role_claim/2/ippo_cer/checkpoints/run_3_seed_2.pkl",
          "checkpoint_sha256": "c4d325f5669be7a051a15436ab1a57c336c1f67f5394aa8d1dfce28cfc645942",
          "run_seed": 2,
          "config": {
            "NUM_ENVS": 1,
            "TOTAL_TIMESTEPS": 200192,
            "NUM_STEPS": 256,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 1.0,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 2,
            "COUNTERFACTUAL_REPLAY": true
          },
          "episodes": 8,
          "diagnostic_seed0": 812020000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed & ~player_1_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 4.865191185667095e-06,
                  "mean_kl_behavior_to_synthetic": -1.3992305850091992e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999996268134,
                    "ess_fraction": 0.99999999981778,
                    "log_weight_quantiles": [
                      -6.532669067382812e-05,
                      -1.3864040374755857e-05,
                      0.0,
                      2.0503997802734375e-05,
                      4.208087921142578e-05
                    ],
                    "log_mean_weight": 1.4729828407834925e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999293296642,
                    "ess_fraction": 0.9999999116620802,
                    "log_weight_quantiles": [
                      -0.00042641162872314453,
                      -0.00033578872680664065,
                      -3.045797348022461e-05,
                      0.00038472414016723635,
                      0.00040900707244873047
                    ],
                    "log_mean_weight": 3.772920624392831e-05
                  }
                },
                "2": {
                  "mean_tv": 1.093991731644303e-06,
                  "mean_kl_behavior_to_synthetic": -5.320562665954349e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.999999921649,
                    "ess_fraction": 0.9999999999617427,
                    "log_weight_quantiles": [
                      -6.270408630371094e-05,
                      -2.384185791015625e-07,
                      0.0,
                      3.814697265625e-06,
                      3.6716461181640625e-05
                    ],
                    "log_mean_weight": -1.703571604035135e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.9999999532471495,
                    "ess_fraction": 0.9999999941558937,
                    "log_weight_quantiles": [
                      -0.00019598007202148438,
                      -7.323026657104492e-05,
                      1.3530254364013672e-05,
                      5.32388687133789e-05,
                      5.924701690673828e-05
                    ],
                    "log_mean_weight": -4.363117905219671e-06
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed",
                      2
                    ],
                    [
                      "player_1_resource_claimed & ~player_0_resource_claimed",
                      1
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 2.5056362567198785e-06,
                  "mean_kl_behavior_to_synthetic": 2.8750686701019816e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.999999528831,
                    "ess_fraction": 0.999999999769937,
                    "log_weight_quantiles": [
                      -0.000194549560546875,
                      -3.349781036376939e-06,
                      0.0,
                      8.106231689453125e-06,
                      0.00012731552124023438
                    ],
                    "log_mean_weight": -4.582702975610819e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999570505016,
                    "ess_fraction": 0.999999946313127,
                    "log_weight_quantiles": [
                      -0.0006306767463684082,
                      -0.00033861398696899414,
                      -0.00010639429092407227,
                      0.00010004639625549314,
                      0.00018996000289916992
                    ],
                    "log_mean_weight": -0.00011731979690501475
                  }
                },
                "2": {
                  "mean_tv": 7.1984859323810725e-06,
                  "mean_kl_behavior_to_synthetic": 5.988707026864038e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999986120033,
                    "ess_fraction": 0.9999999993222672,
                    "log_weight_quantiles": [
                      -0.00037360191345214844,
                      -1.6111135482788083e-05,
                      -1.1920928955078125e-07,
                      2.9158592224121083e-05,
                      0.00018846988677978516
                    ],
                    "log_mean_weight": 5.29621135401747e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999272497992,
                    "ess_fraction": 0.999999909062249,
                    "log_weight_quantiles": [
                      -0.000247955322265625,
                      -0.00015374422073364256,
                      0.00011470913887023926,
                      0.00041213631629943837,
                      0.0007957816123962402
                    ],
                    "log_mean_weight": 0.00013554171888433139
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "gridworlds/territory/role_claim/2/ippo_cer/checkpoints/run_4_seed_3.pkl",
          "checkpoint_sha256": "ee966e36855f5621590301bf39a25889c93efdbb59df916a1c27db265b1cb433",
          "run_seed": 3,
          "config": {
            "NUM_ENVS": 1,
            "TOTAL_TIMESTEPS": 200192,
            "NUM_STEPS": 256,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 1.0,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 3,
            "COUNTERFACTUAL_REPLAY": true
          },
          "episodes": 8,
          "diagnostic_seed0": 812030000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed & ~player_1_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.3462720471015957e-06,
                  "mean_kl_behavior_to_synthetic": 2.1534299166398117e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999999096901,
                    "ess_fraction": 0.9999999999559034,
                    "log_weight_quantiles": [
                      -3.6716461181640625e-05,
                      -2.384185791015625e-07,
                      0.0,
                      5.602836608886719e-06,
                      3.9577484130859375e-05
                    ],
                    "log_mean_weight": 7.257789742133846e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999946389109,
                    "ess_fraction": 0.9999999932986386,
                    "log_weight_quantiles": [
                      -1.7195940017700195e-05,
                      -1.677870750427246e-05,
                      -1.475214958190918e-05,
                      7.172822952270504e-05,
                      0.0002345740795135498
                    ],
                    "log_mean_weight": 1.857764769840314e-05
                  }
                },
                "2": {
                  "mean_tv": 3.973131049232621e-06,
                  "mean_kl_behavior_to_synthetic": 3.5106723586557926e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.999999750735,
                    "ess_fraction": 0.9999999998782886,
                    "log_weight_quantiles": [
                      -5.459785461425781e-05,
                      -7.808208465576172e-06,
                      -4.708766937255859e-06,
                      1.9073486328125e-05,
                      6.198883056640625e-05
                    ],
                    "log_mean_weight": -4.538477597161366e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999447039942,
                    "ess_fraction": 0.9999999308799927,
                    "log_weight_quantiles": [
                      -0.0005088150501251221,
                      -0.0003461778163909912,
                      -2.682209014892578e-07,
                      0.0002754002809524536,
                      0.00035578012466430664
                    ],
                    "log_mean_weight": -1.1599518677023364e-05
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed",
                      2
                    ],
                    [
                      "player_1_resource_claimed & ~player_0_resource_claimed",
                      1
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.5502088422884105e-05,
                  "mean_kl_behavior_to_synthetic": 3.838593908458855e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999937168284,
                    "ess_fraction": 0.9999999969320451,
                    "log_weight_quantiles": [
                      -0.00014030933380126953,
                      -3.9732456207275385e-05,
                      -9.059906005859375e-06,
                      5.2285194396972645e-05,
                      0.00044274330139160156
                    ],
                    "log_mean_weight": -2.0434667607273664e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999997962748194,
                    "ess_fraction": 0.9999997453435242,
                    "log_weight_quantiles": [
                      -0.0008754730224609375,
                      -0.0007938623428344726,
                      1.6987323760986328e-05,
                      0.0005845308303833007,
                      0.0006210803985595703
                    ],
                    "log_mean_weight": -5.2578063825470256e-05
                  }
                },
                "2": {
                  "mean_tv": 3.6377177949957574e-06,
                  "mean_kl_behavior_to_synthetic": 4.558012324926973e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999989792702,
                    "ess_fraction": 0.9999999995015968,
                    "log_weight_quantiles": [
                      -0.00027108192443847656,
                      -2.384185791015625e-07,
                      0.0,
                      1.4781951904296875e-05,
                      7.43865966796875e-05
                    ],
                    "log_mean_weight": 2.0368499642172463e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999998134943564,
                    "ess_fraction": 0.9999997668679454,
                    "log_weight_quantiles": [
                      -0.0008006095886230469,
                      -0.0002824068069458007,
                      2.9385089874267578e-05,
                      0.00036746263504028304,
                      0.0011113882064819336
                    ],
                    "log_mean_weight": 5.2196100764530413e-05
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "gridworlds/territory/role_claim/2/ippo_cer/checkpoints/run_5_seed_4.pkl",
          "checkpoint_sha256": "f7b83a9080ec2350d548b52ec981285b0ce50a270369eac4cc792b4750999329",
          "run_seed": 4,
          "config": {
            "NUM_ENVS": 1,
            "TOTAL_TIMESTEPS": 200192,
            "NUM_STEPS": 256,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 1.0,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 4,
            "COUNTERFACTUAL_REPLAY": true
          },
          "episodes": 8,
          "diagnostic_seed0": 812040000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed & ~player_1_resource_claimed",
                      1
                    ],
                    [
                      "player_1_resource_claimed",
                      2
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.2698020113865368e-06,
                  "mean_kl_behavior_to_synthetic": 3.218575827932016e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999998973078,
                    "ess_fraction": 0.9999999999498573,
                    "log_weight_quantiles": [
                      -9.107589721679688e-05,
                      -2.384185791015625e-07,
                      0.0,
                      5.412101745605466e-06,
                      7.367134094238281e-05
                    ],
                    "log_mean_weight": -4.121505620970599e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999978906897,
                    "ess_fraction": 0.9999999973633621,
                    "log_weight_quantiles": [
                      -0.00011917948722839355,
                      -5.726218223571776e-05,
                      -1.3783574104309082e-05,
                      4.6578049659729e-05,
                      6.833672523498535e-05
                    ],
                    "log_mean_weight": -1.0556154358220038e-05
                  }
                },
                "2": {
                  "mean_tv": 3.13966710346657e-06,
                  "mean_kl_behavior_to_synthetic": 8.443666150869591e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999996814036,
                    "ess_fraction": 0.9999999998444353,
                    "log_weight_quantiles": [
                      -3.838539123535156e-05,
                      -8.860230445861815e-06,
                      -8.940696716308594e-08,
                      1.4448165893554666e-05,
                      0.00013756752014160156
                    ],
                    "log_mean_weight": 5.139932186215242e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999999792968544,
                    "ess_fraction": 0.999999974121068,
                    "log_weight_quantiles": [
                      -6.908178329467773e-05,
                      -3.639161586761474e-05,
                      9.09268856048584e-05,
                      0.0003518998622894287,
                      0.0003534853458404541
                    ],
                    "log_mean_weight": 0.00013157529127984019
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  1
                ],
                "transitions": [
                  [
                    [
                      "player_0_resource_claimed",
                      2
                    ],
                    [
                      "player_1_resource_claimed & ~player_0_resource_claimed",
                      1
                    ],
                    [
                      "~player_0_resource_claimed & ~player_1_resource_claimed",
                      0
                    ]
                  ],
                  [
                    [
                      "True",
                      1
                    ]
                  ],
                  [
                    [
                      "True",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.6666666666666666,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 2.4827816197315685e-05,
                  "mean_kl_behavior_to_synthetic": 1.7297205411411298e-09,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999790168365,
                    "ess_fraction": 0.9999999897543147,
                    "log_weight_quantiles": [
                      -0.0006899833679199219,
                      -7.158517837524413e-05,
                      0.0,
                      6.769895553588866e-05,
                      0.0007652640342712402
                    ],
                    "log_mean_weight": 4.936013043534875e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999967558008241,
                    "ess_fraction": 0.9999959447510302,
                    "log_weight_quantiles": [
                      -0.0031858086585998535,
                      -0.0015942752361297607,
                      2.753734588623047e-05,
                      0.0023373544216156,
                      0.004321753978729248
                    ],
                    "log_mean_weight": 0.00012707657986163468
                  }
                },
                "2": {
                  "mean_tv": 5.251716353827405e-06,
                  "mean_kl_behavior_to_synthetic": -8.075695549439444e-10,
                  "action_weights": {
                    "count": 2048,
                    "ess": 2047.9999939952756,
                    "ess_fraction": 0.9999999970680057,
                    "log_weight_quantiles": [
                      -0.0005049705505371094,
                      -6.341934204101554e-06,
                      0.0,
                      2.640485763549802e-06,
                      0.0018439292907714844
                    ],
                    "log_mean_weight": 1.0116879956237768e-06
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 7.999998188094776,
                    "ess_fraction": 0.999999773511847,
                    "log_weight_quantiles": [
                      -5.650520324707031e-05,
                      -3.660321235656738e-05,
                      3.725290298461914e-05,
                      0.000814467668533325,
                      0.0014234185218811035
                    ],
                    "log_mean_weight": 0.00025873028421377316
                  }
                }
              }
            }
          }
        }
      ],
      "primary": {
        "seeds": [
          {
            "eval_seed0": 503,
            "p_hat": {
              "player_0": 0.204,
              "player_1": 0.759
            },
            "minimum_p_hat": 0.204,
            "sum_p_hat": 0.963,
            "jointly_certified": false
          },
          {
            "eval_seed0": 1000503,
            "p_hat": {
              "player_0": 0.065,
              "player_1": 0.915
            },
            "minimum_p_hat": 0.065,
            "sum_p_hat": 0.98,
            "jointly_certified": false
          },
          {
            "eval_seed0": 2000503,
            "p_hat": {
              "player_0": 0.186,
              "player_1": 0.726
            },
            "minimum_p_hat": 0.186,
            "sum_p_hat": 0.9119999999999999,
            "jointly_certified": false
          },
          {
            "eval_seed0": 3000503,
            "p_hat": {
              "player_0": 0.924,
              "player_1": 0.039
            },
            "minimum_p_hat": 0.039,
            "sum_p_hat": 0.9630000000000001,
            "jointly_certified": false
          },
          {
            "eval_seed0": 4000503,
            "p_hat": {
              "player_0": 0.645,
              "player_1": 0.15
            },
            "minimum_p_hat": 0.15,
            "sum_p_hat": 0.795,
            "jointly_certified": false
          }
        ],
        "jointly_certified_count": 0,
        "mean_minimum_p_hat": 0.1288,
        "mean_sum_p_hat": 0.9226000000000001
      },
      "confirmatory": {
        "seeds": [
          {
            "eval_seed0": 500503,
            "p_hat": {
              "player_0": 0.204,
              "player_1": 0.759
            },
            "minimum_p_hat": 0.204,
            "sum_p_hat": 0.963,
            "jointly_certified": false
          },
          {
            "eval_seed0": 1500503,
            "p_hat": {
              "player_0": 0.064,
              "player_1": 0.916
            },
            "minimum_p_hat": 0.064,
            "sum_p_hat": 0.98,
            "jointly_certified": false
          },
          {
            "eval_seed0": 2500503,
            "p_hat": {
              "player_0": 0.186,
              "player_1": 0.726
            },
            "minimum_p_hat": 0.186,
            "sum_p_hat": 0.9119999999999999,
            "jointly_certified": false
          },
          {
            "eval_seed0": 3500503,
            "p_hat": {
              "player_0": 0.924,
              "player_1": 0.039
            },
            "minimum_p_hat": 0.039,
            "sum_p_hat": 0.9630000000000001,
            "jointly_certified": false
          },
          {
            "eval_seed0": 4500503,
            "p_hat": {
              "player_0": 0.645,
              "player_1": 0.15
            },
            "minimum_p_hat": 0.15,
            "sum_p_hat": 0.795,
            "jointly_certified": false
          }
        ],
        "jointly_certified_count": 0,
        "mean_minimum_p_hat": 0.1286,
        "mean_sum_p_hat": 0.9226000000000001
      }
    },
    "stag_hunt/pr2_ippo": {
      "checkpoints": [
        {
          "checkpoint": "matrix/stag_hunt/sustained_stag/2/pr2_ippo/checkpoints/run_1_seed_0.pkl",
          "checkpoint_sha256": "a6ee7f0c65b13d8ef0ab62db6cc6f18511f5298272a2835b53cc44bf2e31b78c",
          "run_seed": 0,
          "config": {
            "NUM_ENVS": 8,
            "TOTAL_TIMESTEPS": 500224,
            "NUM_STEPS": 32,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 0.99,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 0
          },
          "episodes": 8,
          "diagnostic_seed0": 812000000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.8297472344450083e-06,
                  "mean_kl_behavior_to_synthetic": 1.4892866624270772e-06,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999069055,
                    "ess_fraction": 0.999999999941816,
                    "log_weight_quantiles": [
                      -3.504683877508796e-05,
                      -1.549719200966132e-07,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -1.8297988062276299e-06
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -3.65965579760541e-05,
                      -3.65965579760541e-05,
                      -3.65965579760541e-05,
                      -3.65965579760541e-05,
                      -3.65965579760541e-05
                    ],
                    "log_mean_weight": -3.65965579760541e-05
                  }
                },
                "2": {
                  "mean_tv": 0.004771471408913722,
                  "mean_kl_behavior_to_synthetic": 0.005013527752031263,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.93015581863128,
                    "ess_fraction": 0.9995634738664455,
                    "log_weight_quantiles": [
                      -0.10029349476349125,
                      -1.549719200966132e-07,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -0.0047829027533617605
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -0.10029504448269222,
                      -0.10029504448269222,
                      -0.10029504448269222,
                      -0.10029504448269222,
                      -0.10029504448269222
                    ],
                    "log_mean_weight": -0.10029504448269222
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 0.00024934689020751155,
                  "mean_kl_behavior_to_synthetic": 0.00023517485861736768,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99981229640937,
                    "ess_fraction": 0.9999988268525586,
                    "log_weight_quantiles": [
                      -0.004981753256288357,
                      -1.7881039639178135e-06,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -0.0002493933851327728
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -0.004999634295927535,
                      -0.004999634295927535,
                      -0.004999634295927535,
                      -0.004999634295927535,
                      -0.004999634295927535
                    ],
                    "log_mean_weight": -0.004999634295927535
                  }
                },
                "2": {
                  "mean_tv": 0.04330478586909581,
                  "mean_kl_behavior_to_synthetic": 0.10050865009172383,
                  "action_weights": {
                    "count": 160,
                    "ess": 154.00414126891374,
                    "ess_fraction": 0.9625258829307108,
                    "log_weight_quantiles": [
                      -2.010921718858299,
                      -1.7881039639178135e-06,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -0.0442733942330796
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -2.010939599897938,
                      -2.010939599897938,
                      -2.010939599897938,
                      -2.010939599897938,
                      -2.010939599897938
                    ],
                    "log_mean_weight": -2.010939599897938
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "matrix/stag_hunt/sustained_stag/2/pr2_ippo/checkpoints/run_2_seed_1.pkl",
          "checkpoint_sha256": "1bca01db22814cb97f1486d584bb668f6c729f7caee48d53129e798d606707b7",
          "run_seed": 1,
          "config": {
            "NUM_ENVS": 8,
            "TOTAL_TIMESTEPS": 500224,
            "NUM_STEPS": 32,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 0.99,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 1
          },
          "episodes": 8,
          "diagnostic_seed0": 812010000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 3.814573429650192e-07,
                  "mean_kl_behavior_to_synthetic": 3.8259638082371677e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999955767,
                    "ess_fraction": 0.9999999999972354,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      7.629350875504315e-06
                    ],
                    "log_mean_weight": 3.814689263813925e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      7.629350875504315e-06,
                      7.629350875504315e-06,
                      7.629350875504315e-06,
                      7.629350875504315e-06,
                      7.629350875504315e-06
                    ],
                    "log_mean_weight": 7.629350875504315e-06
                  }
                },
                "2": {
                  "mean_tv": 4.1104814866711195e-07,
                  "mean_kl_behavior_to_synthetic": 5.278074804699674e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999948585,
                    "ess_fraction": 0.9999999999967866,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      8.22539641376352e-06
                    ],
                    "log_mean_weight": 4.112714278524908e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      8.22539641376352e-06,
                      8.22539641376352e-06,
                      8.22539641376352e-06,
                      8.22539641376352e-06,
                      8.22539641376352e-06
                    ],
                    "log_mean_weight": 8.22539641376352e-06
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 5.389479898977702e-08,
                  "mean_kl_behavior_to_synthetic": 3.8303970088416603e-10,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999999307,
                    "ess_fraction": 0.9999999999999567,
                    "log_weight_quantiles": [
                      -9.536488505546004e-07,
                      -1.1920838005607905e-08,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -5.3642839832263264e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -1.0728572306106798e-06,
                      -1.0728572306106798e-06,
                      -1.0728572306106798e-06,
                      -1.0728572306106798e-06,
                      -1.0728572306106798e-06
                    ],
                    "log_mean_weight": -1.0728572306106798e-06
                  }
                },
                "2": {
                  "mean_tv": 5.487817582530833e-07,
                  "mean_kl_behavior_to_synthetic": 8.758036437057164e-08,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999910662,
                    "ess_fraction": 0.9999999999944164,
                    "log_weight_quantiles": [
                      -1.0847701560123824e-05,
                      -1.1920838005607905e-08,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -5.483427051467913e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -1.0966909940179903e-05,
                      -1.0966909940179903e-05,
                      -1.0966909940179903e-05,
                      -1.0966909940179903e-05,
                      -1.0966909940179903e-05
                    ],
                    "log_mean_weight": -1.0966909940179903e-05
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "matrix/stag_hunt/sustained_stag/2/pr2_ippo/checkpoints/run_3_seed_2.pkl",
          "checkpoint_sha256": "efde728ef2eea5537841ac75a8b97afa908438871d68f01a9f19337444891c0b",
          "run_seed": 2,
          "config": {
            "NUM_ENVS": 8,
            "TOTAL_TIMESTEPS": 500224,
            "NUM_STEPS": 32,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 0.99,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 2
          },
          "episodes": 8,
          "diagnostic_seed0": 812020000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 6.027530523196693e-07,
                  "mean_kl_behavior_to_synthetic": 2.2874037706207848e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.9999999988983,
                    "ess_fraction": 0.9999999999931143,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      1.203991450893227e-05
                    ],
                    "log_mean_weight": 6.019991682636228e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      1.203991450893227e-05,
                      1.203991450893227e-05,
                      1.203991450893227e-05,
                      1.203991450893227e-05,
                      1.203991450893227e-05
                    ],
                    "log_mean_weight": 1.203991450893227e-05
                  }
                },
                "2": {
                  "mean_tv": 6.793169651793751e-07,
                  "mean_kl_behavior_to_synthetic": 3.099705162153404e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999859637,
                    "ess_fraction": 0.9999999999912272,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      1.3589617083198391e-05
                    ],
                    "log_mean_weight": 6.794852402409714e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      1.3589617083198391e-05,
                      1.3589617083198391e-05,
                      1.3589617083198391e-05,
                      1.3589617083198391e-05,
                      1.3589617083198391e-05
                    ],
                    "log_mean_weight": 1.3589617083198391e-05
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 0.011672190362221928,
                  "mean_kl_behavior_to_synthetic": 0.013266829095700889,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.5770571513486,
                    "ess_fraction": 0.9973566071959288,
                    "log_weight_quantiles": [
                      -0.26586631134705385,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -0.011741583461119183
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -0.26586631134705385,
                      -0.26586631134705385,
                      -0.26586631134705385,
                      -0.26586631134705385,
                      -0.26586631134705385
                    ],
                    "log_mean_weight": -0.26586631134705385
                  }
                },
                "2": {
                  "mean_tv": 0.04997808880950755,
                  "mean_kl_behavior_to_synthetic": 0.3942524396112438,
                  "action_weights": {
                    "count": 160,
                    "ess": 152.0060135883282,
                    "ess_fraction": 0.9500375849270511,
                    "log_weight_quantiles": [
                      -7.8861442823763355,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -0.0512735095195331
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -7.8861442823763355,
                      -7.8861442823763355,
                      -7.8861442823763355,
                      -7.8861442823763355,
                      -7.8861442823763355
                    ],
                    "log_mean_weight": -7.8861442823763355
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "matrix/stag_hunt/sustained_stag/2/pr2_ippo/checkpoints/run_4_seed_3.pkl",
          "checkpoint_sha256": "e858aca97cb7fbc482b8cd4dab295144dfc4a9429e07221ad05c198496320902",
          "run_seed": 3,
          "config": {
            "NUM_ENVS": 8,
            "TOTAL_TIMESTEPS": 500224,
            "NUM_STEPS": 32,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 0.99,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 3
          },
          "episodes": 8,
          "diagnostic_seed0": 812030000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 3.207810237925069e-07,
                  "mean_kl_behavior_to_synthetic": 1.818790252102111e-08,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.9999999996851,
                    "ess_fraction": 0.9999999999980318,
                    "log_weight_quantiles": [
                      -6.436901458073407e-06,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -3.2184408887959605e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -6.436901458073407e-06,
                      -6.436901458073407e-06,
                      -6.436901458073407e-06,
                      -6.436901458073407e-06,
                      -6.436901458073407e-06
                    ],
                    "log_mean_weight": -6.436901458073407e-06
                  }
                },
                "2": {
                  "mean_tv": 2.1809681642484124e-06,
                  "mean_kl_behavior_to_synthetic": 5.467896316615896e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999998553537,
                    "ess_fraction": 0.9999999999095961,
                    "log_weight_quantiles": [
                      -4.3627063860185444e-05,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -2.181307989632928e-06
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -4.3627063860185444e-05,
                      -4.3627063860185444e-05,
                      -4.3627063860185444e-05,
                      -4.3627063860185444e-05,
                      -4.3627063860185444e-05
                    ],
                    "log_mean_weight": -4.3627063860185444e-05
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 0.046747217379739345,
                  "mean_kl_behavior_to_synthetic": 0.13663099123006012,
                  "action_weights": {
                    "count": 160,
                    "ess": 153.00780528346218,
                    "ess_fraction": 0.9562987830216386,
                    "log_weight_quantiles": [
                      -2.7333163038638304,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -0.04787790395670803
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -2.7333163038638304,
                      -2.7333163038638304,
                      -2.7333163038638304,
                      -2.7333163038638304,
                      -2.7333163038638304
                    ],
                    "log_mean_weight": -2.7333163038638304
                  }
                },
                "2": {
                  "mean_tv": 0.04999642905892633,
                  "mean_kl_behavior_to_synthetic": 0.5535820126968185,
                  "action_weights": {
                    "count": 160,
                    "ess": 152.00024846082985,
                    "ess_fraction": 0.9500015528801866,
                    "log_weight_quantiles": [
                      -11.072806717529602,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -0.05129247707651115
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -11.072806717529602,
                      -11.072806717529602,
                      -11.072806717529602,
                      -11.072806717529602,
                      -11.072806717529602
                    ],
                    "log_mean_weight": -11.072806717529602
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "matrix/stag_hunt/sustained_stag/2/pr2_ippo/checkpoints/run_5_seed_4.pkl",
          "checkpoint_sha256": "0b8d10bc9838666f1863d9c6c72d5bc59cc3e616366ea9c97ebc9a52e5defc94",
          "run_seed": 4,
          "config": {
            "NUM_ENVS": 8,
            "TOTAL_TIMESTEPS": 500224,
            "NUM_STEPS": 32,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 0.99,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 4
          },
          "episodes": 8,
          "diagnostic_seed0": 812040000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.0737913365350134e-07,
                  "mean_kl_behavior_to_synthetic": 5.3562315533616044e-08,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.999999999965,
                    "ess_fraction": 0.9999999999997813,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      2.145761641259014e-06
                    ],
                    "log_mean_weight": 1.0728819142790014e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      2.145761641259014e-06,
                      2.145761641259014e-06,
                      2.145761641259014e-06,
                      2.145761641259014e-06,
                      2.145761641259014e-06
                    ],
                    "log_mean_weight": 2.145761641259014e-06
                  }
                },
                "2": {
                  "mean_tv": 1.234745864594496e-07,
                  "mean_kl_behavior_to_synthetic": 7.366259157362131e-08,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999995237,
                    "ess_fraction": 0.9999999999997022,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      2.5033889414771693e-06
                    ],
                    "log_mean_weight": 1.251695959090595e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      2.5033889414771693e-06,
                      2.5033889414771693e-06,
                      2.5033889414771693e-06,
                      2.5033889414771693e-06,
                      2.5033889414771693e-06
                    ],
                    "log_mean_weight": 2.5033889414771693e-06
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "canonical",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 2.3062393312123593e-05,
                  "mean_kl_behavior_to_synthetic": 2.0078113025840938e-05,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.999998476205,
                    "ess_fraction": 0.9999999904762813,
                    "log_weight_quantiles": [
                      -0.0004483575321501121,
                      -1.299368386753481e-06,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -2.3062797663293836e-05
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -0.00046135121601764695,
                      -0.00046135121601764695,
                      -0.00046135121601764695,
                      -0.00046135121601764695,
                      -0.00046135121601764695
                    ],
                    "log_mean_weight": -0.00046135121601764695
                  }
                },
                "2": {
                  "mean_tv": 0.01908399491014204,
                  "mean_kl_behavior_to_synthetic": 0.02402872912792873,
                  "action_weights": {
                    "count": 160,
                    "ess": 158.85759610820836,
                    "ess_fraction": 0.9928599756763022,
                    "log_weight_quantiles": [
                      -0.4807379247067729,
                      -1.299368386753481e-06,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -0.019268758324183944
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -0.48075091839064044,
                      -0.48075091839064044,
                      -0.48075091839064044,
                      -0.48075091839064044,
                      -0.48075091839064044
                    ],
                    "log_mean_weight": -0.48075091839064044
                  }
                }
              }
            }
          }
        }
      ],
      "primary": {
        "seeds": [
          {
            "eval_seed0": 503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          },
          {
            "eval_seed0": 1000503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          },
          {
            "eval_seed0": 2000503,
            "p_hat": {
              "player_0": 0.0,
              "player_1": 0.0
            },
            "minimum_p_hat": 0.0,
            "sum_p_hat": 0.0,
            "jointly_certified": false
          },
          {
            "eval_seed0": 3000503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          },
          {
            "eval_seed0": 4000503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          }
        ],
        "jointly_certified_count": 4,
        "mean_minimum_p_hat": 0.8,
        "mean_sum_p_hat": 1.6
      },
      "confirmatory": {
        "seeds": [
          {
            "eval_seed0": 500503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          },
          {
            "eval_seed0": 1500503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          },
          {
            "eval_seed0": 2500503,
            "p_hat": {
              "player_0": 0.0,
              "player_1": 0.0
            },
            "minimum_p_hat": 0.0,
            "sum_p_hat": 0.0,
            "jointly_certified": false
          },
          {
            "eval_seed0": 3500503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          },
          {
            "eval_seed0": 4500503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          }
        ],
        "jointly_certified_count": 4,
        "mean_minimum_p_hat": 0.8,
        "mean_sum_p_hat": 1.6
      }
    },
    "stag_hunt/pr2_ippo_cer": {
      "checkpoints": [
        {
          "checkpoint": "matrix/stag_hunt/sustained_stag/2/pr2_ippo_cer/checkpoints/run_1_seed_0.pkl",
          "checkpoint_sha256": "bcbf9a8d341a37f069e3aaa6b0033e71868429972843013762cd15b9bb70527e",
          "run_seed": 0,
          "config": {
            "NUM_ENVS": 8,
            "TOTAL_TIMESTEPS": 500224,
            "NUM_STEPS": 32,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 0.99,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 0,
            "COUNTERFACTUAL_REPLAY": true
          },
          "episodes": 8,
          "diagnostic_seed0": 812000000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 6.7033302702027614e-09,
                  "mean_kl_behavior_to_synthetic": 5.947499432485936e-09,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.9999999999999,
                    "ess_fraction": 0.9999999999999993,
                    "log_weight_quantiles": [
                      -1.1920928244535389e-07,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -5.960463806964352e-09
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -1.1920928244535389e-07,
                      -1.1920928244535389e-07,
                      -1.1920928244535389e-07,
                      -1.1920928244535389e-07,
                      -1.1920928244535389e-07
                    ],
                    "log_mean_weight": -1.1920928244535389e-07
                  }
                },
                "2": {
                  "mean_tv": 6.705721852858923e-09,
                  "mean_kl_behavior_to_synthetic": 5.947435546939987e-09,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.9999999999999,
                    "ess_fraction": 0.9999999999999993,
                    "log_weight_quantiles": [
                      -1.1920928244535389e-07,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -5.960463806964352e-09
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -1.1920928244535389e-07,
                      -1.1920928244535389e-07,
                      -1.1920928244535389e-07,
                      -1.1920928244535389e-07,
                      -1.1920928244535389e-07
                    ],
                    "log_mean_weight": -1.1920928244535389e-07
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.3366060251711047e-08,
                  "mean_kl_behavior_to_synthetic": 2.4308729930307425e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999999957,
                    "ess_fraction": 0.9999999999999973,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      2.3841855067985307e-07
                    ],
                    "log_mean_weight": 1.192092890623668e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      2.3841855067985307e-07,
                      2.3841855067985307e-07,
                      2.3841855067985307e-07,
                      2.3841855067985307e-07,
                      2.3841855067985307e-07
                    ],
                    "log_mean_weight": 2.3841855067985307e-07
                  }
                },
                "2": {
                  "mean_tv": 1.9179569177067286e-08,
                  "mean_kl_behavior_to_synthetic": 2.490477634222814e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999999943,
                    "ess_fraction": 0.9999999999999964,
                    "log_weight_quantiles": [
                      -1.1920928244535389e-07,
                      0.0,
                      0.0,
                      0.0,
                      2.3841855067985307e-07
                    ],
                    "log_mean_weight": 5.960465192531569e-09
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      1.1920926823449918e-07,
                      1.1920926823449918e-07,
                      1.1920926823449918e-07,
                      1.1920926823449918e-07,
                      1.1920926823449918e-07
                    ],
                    "log_mean_weight": 1.1920926823449918e-07
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "matrix/stag_hunt/sustained_stag/2/pr2_ippo_cer/checkpoints/run_2_seed_1.pkl",
          "checkpoint_sha256": "c346f5aee993b476fad717e5942d85174d64a70a461186813634e111e6a81e56",
          "run_seed": 1,
          "config": {
            "NUM_ENVS": 8,
            "TOTAL_TIMESTEPS": 500224,
            "NUM_STEPS": 32,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 0.99,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 1,
            "COUNTERFACTUAL_REPLAY": true
          },
          "episodes": 8,
          "diagnostic_seed0": 812010000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 4.209450675080677e-08,
                  "mean_kl_behavior_to_synthetic": 9.751162546937976e-09,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.9999999999947,
                    "ess_fraction": 0.9999999999999669,
                    "log_weight_quantiles": [
                      -8.344636626134161e-07,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -4.172316657058062e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -8.344636626134161e-07,
                      -8.344636626134161e-07,
                      -8.344636626134161e-07,
                      -8.344636626134161e-07,
                      -8.344636626134161e-07
                    ],
                    "log_mean_weight": -8.344636626134161e-07
                  }
                },
                "2": {
                  "mean_tv": 4.1644710573123257e-07,
                  "mean_kl_behavior_to_synthetic": 2.943761299135817e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.9999999994708,
                    "ess_fraction": 0.9999999999966924,
                    "log_weight_quantiles": [
                      -8.344605475940625e-06,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -4.1722861998497274e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -8.344605475940625e-06,
                      -8.344605475940625e-06,
                      -8.344605475940625e-06,
                      -8.344605475940625e-06,
                      -8.344605475940625e-06
                    ],
                    "log_mean_weight": -8.344605475940625e-06
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.5472358626024434e-07,
                  "mean_kl_behavior_to_synthetic": 3.1542645319615305e-08,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999992698,
                    "ess_fraction": 0.9999999999995437,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      3.0994169719633646e-06
                    ],
                    "log_mean_weight": 1.549710768492618e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      3.0994169719633646e-06,
                      3.0994169719633646e-06,
                      3.0994169719633646e-06,
                      3.0994169719633646e-06,
                      3.0994169719633646e-06
                    ],
                    "log_mean_weight": 3.0994169719633646e-06
                  }
                },
                "2": {
                  "mean_tv": 1.8467416831220277e-07,
                  "mean_kl_behavior_to_synthetic": 4.837438690531366e-08,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.9999999998962,
                    "ess_fraction": 0.9999999999993513,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      3.695459781738464e-06
                    ],
                    "log_mean_weight": 1.847733134898925e-07
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      3.695459781738464e-06,
                      3.695459781738464e-06,
                      3.695459781738464e-06,
                      3.695459781738464e-06,
                      3.695459781738464e-06
                    ],
                    "log_mean_weight": 3.695459781738464e-06
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "matrix/stag_hunt/sustained_stag/2/pr2_ippo_cer/checkpoints/run_3_seed_2.pkl",
          "checkpoint_sha256": "cad01c8f6a6b2dd1daea18477da3017f082bb43994773279834e8599833ad374",
          "run_seed": 2,
          "config": {
            "NUM_ENVS": 8,
            "TOTAL_TIMESTEPS": 500224,
            "NUM_STEPS": 32,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 0.99,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 2,
            "COUNTERFACTUAL_REPLAY": true
          },
          "episodes": 8,
          "diagnostic_seed0": 812020000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 3.0761750568843286e-48,
                  "mean_kl_behavior_to_synthetic": -6.225174216698491e-53,
                  "action_weights": {
                    "count": 160,
                    "ess": 160.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": 0.0
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": 0.0
                  }
                },
                "2": {
                  "mean_tv": 7.793311164823329e-38,
                  "mean_kl_behavior_to_synthetic": -1.6763464963355024e-52,
                  "action_weights": {
                    "count": 160,
                    "ess": 160.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": 0.0
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": 0.0
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 2.3918019050979186e-08,
                  "mean_kl_behavior_to_synthetic": 9.816661099055026e-08,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999999827,
                    "ess_fraction": 0.9999999999999891,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      4.768370445162873e-07
                    ],
                    "log_mean_weight": 2.384185758159706e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      4.768370445162873e-07,
                      4.768370445162873e-07,
                      4.768370445162873e-07,
                      4.768370445162873e-07,
                      4.768370445162873e-07
                    ],
                    "log_mean_weight": 4.768370445162873e-07
                  }
                },
                "2": {
                  "mean_tv": 2.3941068508157688e-08,
                  "mean_kl_behavior_to_synthetic": 1.0673013970558656e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999999827,
                    "ess_fraction": 0.9999999999999891,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      4.768370445162873e-07
                    ],
                    "log_mean_weight": 2.384185758159706e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      4.768370445162873e-07,
                      4.768370445162873e-07,
                      4.768370445162873e-07,
                      4.768370445162873e-07,
                      4.768370445162873e-07
                    ],
                    "log_mean_weight": 4.768370445162873e-07
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "matrix/stag_hunt/sustained_stag/2/pr2_ippo_cer/checkpoints/run_4_seed_3.pkl",
          "checkpoint_sha256": "1734436c7fbba046ef6094ec4aee8db23529b979ba5c1ea557c3e3e88667dd5c",
          "run_seed": 3,
          "config": {
            "NUM_ENVS": 8,
            "TOTAL_TIMESTEPS": 500224,
            "NUM_STEPS": 32,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 0.99,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 3,
            "COUNTERFACTUAL_REPLAY": true
          },
          "episodes": 8,
          "diagnostic_seed0": 812030000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 1.0547391808837163e-06,
                  "mean_kl_behavior_to_synthetic": 2.8243159532752804e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.9999999966165,
                    "ess_fraction": 0.9999999999788531,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      2.109913475578651e-05
                    ],
                    "log_mean_weight": 1.054967310504615e-06
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      2.109913475578651e-05,
                      2.109913475578651e-05,
                      2.109913475578651e-05,
                      2.109913475578651e-05,
                      2.109913475578651e-05
                    ],
                    "log_mean_weight": 2.109913475578651e-05
                  }
                },
                "2": {
                  "mean_tv": 1.2410815247884881e-06,
                  "mean_kl_behavior_to_synthetic": 4.260238589102799e-07,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999532756,
                    "ess_fraction": 0.9999999999707972,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      0.0,
                      2.4794508135528304e-05
                    ],
                    "log_mean_weight": 1.2397400073739135e-06
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      2.4794508135528304e-05,
                      2.4794508135528304e-05,
                      2.4794508135528304e-05,
                      2.4794508135528304e-05,
                      2.4794508135528304e-05
                    ],
                    "log_mean_weight": 2.4794508135528304e-05
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 0.05000000118939581,
                  "mean_kl_behavior_to_synthetic": 42.106005859375,
                  "action_weights": {
                    "count": 160,
                    "ess": 152.0,
                    "ess_fraction": 0.95,
                    "log_weight_quantiles": [
                      -842.1201171875,
                      0.0,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -0.05129329438755058
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -842.1201171875,
                      -842.1201171875,
                      -842.1201171875,
                      -842.1201171875,
                      -842.1201171875
                    ],
                    "log_mean_weight": -842.1201171875
                  }
                },
                "2": {
                  "mean_tv": 0.1,
                  "mean_kl_behavior_to_synthetic": 84.21201171875,
                  "action_weights": {
                    "count": 160,
                    "ess": 144.0,
                    "ess_fraction": 0.9,
                    "log_weight_quantiles": [
                      -842.1201171875,
                      -84.21201171874971,
                      0.0,
                      0.0,
                      0.0
                    ],
                    "log_mean_weight": -0.10536051565782628
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      -1684.240234375,
                      -1684.240234375,
                      -1684.240234375,
                      -1684.240234375,
                      -1684.240234375
                    ],
                    "log_mean_weight": -1684.240234375
                  }
                }
              }
            }
          }
        },
        {
          "checkpoint": "matrix/stag_hunt/sustained_stag/2/pr2_ippo_cer/checkpoints/run_5_seed_4.pkl",
          "checkpoint_sha256": "7179a949b0b47509cc0b2821c7bbfc197390f2a8dfa18257809069b648d51229",
          "run_seed": 4,
          "config": {
            "NUM_ENVS": 8,
            "TOTAL_TIMESTEPS": 500224,
            "NUM_STEPS": 32,
            "NUM_MINIBATCHES": 4,
            "UPDATE_EPOCHS": 4,
            "LR": 0.0003,
            "ANNEAL_LR": false,
            "MAX_GRAD_NORM": 0.5,
            "ACTIVATION": "tanh",
            "GAMMA": 0.99,
            "GAE_LAMBDA": 0.95,
            "CLIP_EPS": 0.2,
            "VF_COEF": 0.5,
            "ENT_COEF": 0.01,
            "SEED": 4,
            "COUNTERFACTUAL_REPLAY": true
          },
          "episodes": 8,
          "diagnostic_seed0": 812040000,
          "per_agent": {
            "player_0": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 7.680151087441592e-08,
                  "mean_kl_behavior_to_synthetic": 2.2845377981445735e-08,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999998448,
                    "ess_fraction": 0.999999999999903,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      1.1920917586393675e-08,
                      1.430507836630568e-06
                    ],
                    "log_mean_weight": 7.748589918658684e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      1.5497170124945114e-06,
                      1.5497170124945114e-06,
                      1.5497170124945114e-06,
                      1.5497170124945114e-06,
                      1.5497170124945114e-06
                    ],
                    "log_mean_weight": 1.5497170124945114e-06
                  }
                },
                "2": {
                  "mean_tv": 8.718398037456271e-08,
                  "mean_kl_behavior_to_synthetic": 2.7671798256501554e-08,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999999997888,
                    "ess_fraction": 0.999999999999868,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      1.1920917586393675e-08,
                      1.6689259609847795e-06
                    ],
                    "log_mean_weight": 8.94068228409861e-08
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      1.788135136848723e-06,
                      1.788135136848723e-06,
                      1.788135136848723e-06,
                      1.788135136848723e-06,
                      1.788135136848723e-06
                    ],
                    "log_mean_weight": 1.788135136848723e-06
                  }
                }
              }
            },
            "player_1": {
              "monitor_numbering": {
                "numbering": "permutation",
                "initial_id": 0,
                "accepting_ids": [
                  2
                ],
                "transitions": [
                  [
                    [
                      "both_stag",
                      1
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "both_stag & ~unmirrored",
                      2
                    ],
                    [
                      "~both_stag",
                      0
                    ]
                  ],
                  [
                    [
                      "both_stag & unmirrored",
                      1
                    ],
                    [
                      "unmirrored & ~both_stag",
                      0
                    ],
                    [
                      "~unmirrored",
                      2
                    ]
                  ]
                ]
              },
              "factual_sample_fraction": 0.3333333333333333,
              "synthetic_sink_sample_fraction": 0.0,
              "by_synthetic_start_state": {
                "1": {
                  "mean_tv": 4.197736766381077e-06,
                  "mean_kl_behavior_to_synthetic": 2.519032824722477e-06,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.9999999477435,
                    "ess_fraction": 0.9999999996733969,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      9.536342986393178e-08,
                      8.296241867356002e-05
                    ],
                    "log_mean_weight": 4.195965942211549e-06
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      8.39160529721994e-05,
                      8.39160529721994e-05,
                      8.39160529721994e-05,
                      8.39160529721994e-05,
                      8.39160529721994e-05
                    ],
                    "log_mean_weight": 8.39160529721994e-05
                  }
                },
                "2": {
                  "mean_tv": 4.428801540068299e-06,
                  "mean_kl_behavior_to_synthetic": 2.968197774552264e-06,
                  "action_weights": {
                    "count": 160,
                    "ess": 159.99999994171955,
                    "ess_fraction": 0.9999999996357471,
                    "log_weight_quantiles": [
                      0.0,
                      0.0,
                      0.0,
                      9.536342986393178e-08,
                      8.76113772392273e-05
                    ],
                    "log_mean_weight": 4.42843269350206e-06
                  },
                  "episode_weights": {
                    "count": 8,
                    "ess": 8.0,
                    "ess_fraction": 1.0,
                    "log_weight_quantiles": [
                      8.856501153786667e-05,
                      8.856501153786667e-05,
                      8.856501153786667e-05,
                      8.856501153786667e-05,
                      8.856501153786667e-05
                    ],
                    "log_mean_weight": 8.856501153786667e-05
                  }
                }
              }
            }
          }
        }
      ],
      "primary": {
        "seeds": [
          {
            "eval_seed0": 503,
            "p_hat": {
              "player_0": 0.0,
              "player_1": 0.0
            },
            "minimum_p_hat": 0.0,
            "sum_p_hat": 0.0,
            "jointly_certified": false
          },
          {
            "eval_seed0": 1000503,
            "p_hat": {
              "player_0": 0.0,
              "player_1": 0.0
            },
            "minimum_p_hat": 0.0,
            "sum_p_hat": 0.0,
            "jointly_certified": false
          },
          {
            "eval_seed0": 2000503,
            "p_hat": {
              "player_0": 0.0,
              "player_1": 0.0
            },
            "minimum_p_hat": 0.0,
            "sum_p_hat": 0.0,
            "jointly_certified": false
          },
          {
            "eval_seed0": 3000503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          },
          {
            "eval_seed0": 4000503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          }
        ],
        "jointly_certified_count": 2,
        "mean_minimum_p_hat": 0.4,
        "mean_sum_p_hat": 0.8
      },
      "confirmatory": {
        "seeds": [
          {
            "eval_seed0": 500503,
            "p_hat": {
              "player_0": 0.0,
              "player_1": 0.0
            },
            "minimum_p_hat": 0.0,
            "sum_p_hat": 0.0,
            "jointly_certified": false
          },
          {
            "eval_seed0": 1500503,
            "p_hat": {
              "player_0": 0.0,
              "player_1": 0.0
            },
            "minimum_p_hat": 0.0,
            "sum_p_hat": 0.0,
            "jointly_certified": false
          },
          {
            "eval_seed0": 2500503,
            "p_hat": {
              "player_0": 0.0,
              "player_1": 0.0
            },
            "minimum_p_hat": 0.0,
            "sum_p_hat": 0.0,
            "jointly_certified": false
          },
          {
            "eval_seed0": 3500503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          },
          {
            "eval_seed0": 4500503,
            "p_hat": {
              "player_0": 1.0,
              "player_1": 1.0
            },
            "minimum_p_hat": 1.0,
            "sum_p_hat": 2.0,
            "jointly_certified": true
          }
        ],
        "jointly_certified_count": 2,
        "mean_minimum_p_hat": 0.4,
        "mean_sum_p_hat": 0.8
      }
    }
  },
  "scope": "Frozen final checkpoints; stochastic factual rollouts and whole-episode relabelling. Not a training ablation or an unbiased estimator of the original task gradient. Empirical ESS can miss rare high-weight paths.",
  "provenance": {
    "investigation_date": "2026-10-12",
    "analysis_base_commit": "692275686b6d4cd8c4ce4a3571fdcc9e03858c20",
    "source_working_copy": "/vol/gpudata/oja24-pltlf/exports",
    "checkpoint_source": "Temporary copies of trusted project archive checkpoints; inputs unchanged.",
    "diagnostic_protocol": "Eight complete stochastic episodes per checkpoint; recovered legacy monitor numbering; no policy training."
  }
}