{
  "generated_at_utc": "2026-10-14T10:48:49.313900+00:00",
  "analysis_sha256": "0cab13f81d93e0cf3115ebed1842a0892a6dccb993c75ef18c17a3a217f2296e",
  "previous_report_sha256": "2ebd4b843ffba3e838f6699aee36e91c8e589f26d0766760ba1d56dfde67d95d",
  "previous_report": "territory-cer-responses-2026-10-14.json",
  "task": "territory",
  "algorithm": "ippo_cer",
  "comparison_family_size": 35,
  "family_components": {
    "ordinary_ippo_profile_budget_cases": 10,
    "conservative_cer_profile_budget_cases": 20,
    "fixed_existing_warm_responses": 5
  },
  "delta": 0.025,
  "epsilon": 0.05,
  "eval_episodes": 8192,
  "independent_binomial_tail_inversions": 32,
  "profiles": [
    {
      "seed": 1,
      "source_revision": "612145f915d524d75c600f14db763544272e1d83",
      "checkpoint_sha256": "88ecea3936ed6fdfe4fc41bef6f1580aa8660a5274810bcd4ffe217cbcb31d7f",
      "result_sha256": "41dc6308f5881e57ba0099bcb3c19aedd252f223155439039bfa20d9f6546f5c",
      "protocol_sha256": "931fbb3aa065f527bc87d7e5f1a036c9a885e26b1fb9585139480929ceb2bf50",
      "budgets": [
        {
          "budget": 50000,
          "selection": {
            "player_0": {
              "validation_gain": 0.125,
              "candidate": "player_0_ippo_fresh_1_50000"
            },
            "player_1": {
              "validation_gain": 0.0,
              "candidate": "original"
            }
          },
          "gains": {
            "player_0": 0.08935546875,
            "player_1": 0.0
          },
          "paired_counts": {
            "player_0": {
              "positive": 889,
              "negative": 157
            },
            "player_1": {
              "positive": 0,
              "negative": 0
            }
          },
          "assessment": {
            "method": "heldout_searched_deviations_exact_paired_binomial",
            "epsilon": 0.05,
            "delta": 0.025,
            "family_size": 35,
            "max_found_gain": 0.08935546875,
            "max_found_gain_ci": [
              0.0705563184531562,
              0.10793295682948235
            ],
            "per_agent_gain_ci": {
              "player_0": [
                0.0705563184531562,
                0.10793295682948235
              ],
              "player_1": [
                -0.0011374957561519627,
                0.0011374957561519627
              ]
            },
            "empirical_stability_passes": false,
            "profitable_deviation_witness": true,
            "inconclusive": false,
            "epsilon_ne_certified": false,
            "confidence_scope": "rollout_error_for_selected_candidates_only"
          },
          "raw_array_sha256": {
            "player_0": "2cb24aeba0c516dd6d6c6ae36e193c83cc728ebaed6260b1f29ffb55575e2197",
            "player_1": "309ece0294e05da87734c9ecdeef8a1bde4498f38e07c8bbfdb8afbc47dc76fa"
          }
        },
        {
          "budget": 200000,
          "selection": {
            "player_0": {
              "validation_gain": 0.1328125,
              "candidate": "player_0_ippo_fresh_2_200000"
            },
            "player_1": {
              "validation_gain": 0.0625,
              "candidate": "player_1_ippo_warm_0_200000"
            }
          },
          "gains": {
            "player_0": 0.119384765625,
            "player_1": 0.054931640625
          },
          "paired_counts": {
            "player_0": {
              "positive": 1165,
              "negative": 187
            },
            "player_1": {
              "positive": 509,
              "negative": 59
            }
          },
          "assessment": {
            "method": "heldout_searched_deviations_exact_paired_binomial",
            "epsilon": 0.05,
            "delta": 0.025,
            "family_size": 35,
            "max_found_gain": 0.119384765625,
            "max_found_gain_ci": [
              0.09846372076457671,
              0.1400133289277195
            ],
            "per_agent_gain_ci": {
              "player_0": [
                0.09846372076457671,
                0.1400133289277195
              ],
              "player_1": [
                0.04121309525569877,
                0.06850318267392716
              ]
            },
            "empirical_stability_passes": false,
            "profitable_deviation_witness": true,
            "inconclusive": false,
            "epsilon_ne_certified": false,
            "confidence_scope": "rollout_error_for_selected_candidates_only"
          },
          "raw_array_sha256": {
            "player_0": "984070b10f1b305e78b57a97af5b6e376f4113304530a383e445b131f9623eb9",
            "player_1": "5314e719547ca5e192cbd29bd0405bfaf55721e266853c4696599ee99fa776c4"
          }
        }
      ]
    },
    {
      "seed": 3,
      "source_revision": "612145f915d524d75c600f14db763544272e1d83",
      "checkpoint_sha256": "ee966e36855f5621590301bf39a25889c93efdbb59df916a1c27db265b1cb433",
      "result_sha256": "697304ab64f3393df57695d2ee866e68f2e6c4513b3861fb184b3891f23c2564",
      "protocol_sha256": "397ea176b0963155c7fba915a024b022b709c13018d1291c802b670feb6dcaeb",
      "budgets": [
        {
          "budget": 50000,
          "selection": {
            "player_0": {
              "validation_gain": 0.013671875,
              "candidate": "player_0_ippo_warm_0_50000"
            },
            "player_1": {
              "validation_gain": 0.0,
              "candidate": "original"
            }
          },
          "gains": {
            "player_0": 0.023681640625,
            "player_1": 0.0
          },
          "paired_counts": {
            "player_0": {
              "positive": 205,
              "negative": 11
            },
            "player_1": {
              "positive": 0,
              "negative": 0
            }
          },
          "assessment": {
            "method": "heldout_searched_deviations_exact_paired_binomial",
            "epsilon": 0.05,
            "delta": 0.025,
            "family_size": 35,
            "max_found_gain": 0.023681640625,
            "max_found_gain_ci": [
              0.015460176841203595,
              0.03181332713507675
            ],
            "per_agent_gain_ci": {
              "player_0": [
                0.015460176841203595,
                0.03181332713507675
              ],
              "player_1": [
                -0.0011374957561519627,
                0.0011374957561519627
              ]
            },
            "empirical_stability_passes": true,
            "profitable_deviation_witness": false,
            "inconclusive": false,
            "epsilon_ne_certified": false,
            "confidence_scope": "rollout_error_for_selected_candidates_only"
          },
          "raw_array_sha256": {
            "player_0": "d1881e00a39e1c02a21a457f5adcf44c9b90f825d35cbae90f24ec6236a6db7e",
            "player_1": "53244dda1bc683589f7f170ca9cc93f9234aea527c8bf6a3c6ab02af37121184"
          }
        },
        {
          "budget": 200000,
          "selection": {
            "player_0": {
              "validation_gain": 0.013671875,
              "candidate": "player_0_ippo_warm_0_50000"
            },
            "player_1": {
              "validation_gain": 0.00390625,
              "candidate": "player_1_iql_fresh_1_200000"
            }
          },
          "gains": {
            "player_0": 0.023681640625,
            "player_1": 0.0042724609375
          },
          "paired_counts": {
            "player_0": {
              "positive": 205,
              "negative": 11
            },
            "player_1": {
              "positive": 275,
              "negative": 240
            }
          },
          "assessment": {
            "method": "heldout_searched_deviations_exact_paired_binomial",
            "epsilon": 0.05,
            "delta": 0.025,
            "family_size": 35,
            "max_found_gain": 0.023681640625,
            "max_found_gain_ci": [
              0.015460176841203595,
              0.03181332713507675
            ],
            "per_agent_gain_ci": {
              "player_0": [
                0.015460176841203595,
                0.03181332713507675
              ],
              "player_1": [
                -0.010311254730317802,
                0.018845161971138117
              ]
            },
            "empirical_stability_passes": true,
            "profitable_deviation_witness": false,
            "inconclusive": false,
            "epsilon_ne_certified": false,
            "confidence_scope": "rollout_error_for_selected_candidates_only"
          },
          "raw_array_sha256": {
            "player_0": "d1881e00a39e1c02a21a457f5adcf44c9b90f825d35cbae90f24ec6236a6db7e",
            "player_1": "14a1cd9f25be7ddc9c6853d4d3a41e284acdb2129de1d97281e59159dd10a8e9"
          }
        }
      ]
    }
  ],
  "raw_artifact_sha256": {
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_1/result.json": "41dc6308f5881e57ba0099bcb3c19aedd252f223155439039bfa20d9f6546f5c",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_1/protocol.json": "931fbb3aa065f527bc87d7e5f1a036c9a885e26b1fb9585139480929ceb2bf50",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_1/heldout_50000_player_0.npz": "2cb24aeba0c516dd6d6c6ae36e193c83cc728ebaed6260b1f29ffb55575e2197",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_1/heldout_50000_player_1.npz": "309ece0294e05da87734c9ecdeef8a1bde4498f38e07c8bbfdb8afbc47dc76fa",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_1/heldout_200000_player_0.npz": "984070b10f1b305e78b57a97af5b6e376f4113304530a383e445b131f9623eb9",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_1/heldout_200000_player_1.npz": "5314e719547ca5e192cbd29bd0405bfaf55721e266853c4696599ee99fa776c4",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_3/result.json": "697304ab64f3393df57695d2ee866e68f2e6c4513b3861fb184b3891f23c2564",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_3/protocol.json": "397ea176b0963155c7fba915a024b022b709c13018d1291c802b670feb6dcaeb",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_3/heldout_50000_player_0.npz": "d1881e00a39e1c02a21a457f5adcf44c9b90f825d35cbae90f24ec6236a6db7e",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_3/heldout_50000_player_1.npz": "53244dda1bc683589f7f170ca9cc93f9234aea527c8bf6a3c6ab02af37121184",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_3/heldout_200000_player_0.npz": "d1881e00a39e1c02a21a457f5adcf44c9b90f825d35cbae90f24ec6236a6db7e",
    "/var/home/omar/Projects/quartz/pltlf/exports/optimisations/pursuit-rational-refinement-20261012/deviation_candidates/territory/ippo_cer/seed_3/heldout_200000_player_1.npz": "14a1cd9f25be7ddc9c6853d4d3a41e284acdb2129de1d97281e59159dd10a8e9",
    "/tmp/pltlf-coins-full-observation-20261012/docs/AI/experiments/results/territory-cer-responses-2026-10-14.json": "2ebd4b843ffba3e838f6699aee36e91c8e589f26d0766760ba1d56dfde67d95d"
  },
  "scope": "Reassessment of unchanged held-out outcomes under the combined Territory response envelope. Seed one remains a profitable-deviation witness; seed three remains an uncertified finite search. No new training, evaluation, selection, or equilibrium certificate."
}
