{ "observed_at_utc": "2026-10-13T07:58:31.657157+00:00", "source_revision": "612145f915d524d75c600f14db763544272e1d83", "scope": "Complete targeted full-budget 50 profiles; never pooled with half-budget screens.", "statistical_scope": "Task-level Hoeffding satisfaction summaries as originally reported; new exact payoff-ceiling bounds cover all 155 planned CER-study profiles at delta .05. Broader paper selection requires recomputation.", "snapshot_sha256": "da6588c475123fc1d2aa515eee1c586539c1bcd739b91f9e4ef358eaa1cf58bb", "groups": { "stag_hunt/pr2_ippo/cer": { "seeds": [ 100, 101, 102, 103, 104 ], "mean_minimum_satisfaction": 0.1998, "mean_sum_satisfaction": 0.3996, "joint_satisfaction_certified": 1, "joint_rational_synthesis_certified": 1, "exact_joint_satisfaction_certified": 1, "mean_training_seconds": 2573.1733624473213 }, "stag_hunt/pr2_ippo/corrected": { "seeds": [ 100, 101, 102, 103, 104 ], "mean_minimum_satisfaction": 0.4, "mean_sum_satisfaction": 0.8, "joint_satisfaction_certified": 2, "joint_rational_synthesis_certified": 2, "exact_joint_satisfaction_certified": 2, "mean_training_seconds": 2325.964334319532 }, "stag_hunt/pr2_ippo/critic_only": { "seeds": [ 100, 101, 102, 103, 104 ], "mean_minimum_satisfaction": 0.4, "mean_sum_satisfaction": 0.8, "joint_satisfaction_certified": 2, "joint_rational_synthesis_certified": 2, "exact_joint_satisfaction_certified": 2, "mean_training_seconds": 2594.423968476057 }, "stag_hunt/pr2_ippo/no_cer": { "seeds": [ 100, 101, 102, 103, 104 ], "mean_minimum_satisfaction": 0.4048, "mean_sum_satisfaction": 0.8096, "joint_satisfaction_certified": 2, "joint_rational_synthesis_certified": 2, "exact_joint_satisfaction_certified": 2, "mean_training_seconds": 1638.8361788108946 }, "stag_hunt/pr2_ippo/rm_shaping": { "seeds": [ 100, 101, 102, 103, 104 ], "mean_minimum_satisfaction": 1.0, "mean_sum_satisfaction": 2.0, "joint_satisfaction_certified": 5, "joint_rational_synthesis_certified": 5, "exact_joint_satisfaction_certified": 5, "mean_training_seconds": 1495.1121991023422 }, "territory/ippo/cer": { "seeds": [ 100, 101, 102, 103, 104 ], "mean_minimum_satisfaction": 0.1418, "mean_sum_satisfaction": 0.9152, "joint_satisfaction_certified": 1, "joint_rational_synthesis_certified": 0, "exact_joint_satisfaction_certified": 1, "mean_training_seconds": 5883.877221344691 }, "territory/ippo/corrected": { "seeds": [ 100, 101, 102, 103, 104 ], "mean_minimum_satisfaction": 0.13720000000000002, "mean_sum_satisfaction": 0.9092, "joint_satisfaction_certified": 0, "joint_rational_synthesis_certified": 0, "exact_joint_satisfaction_certified": 0, "mean_training_seconds": 6370.437258936092 }, "territory/ippo/critic_only": { "seeds": [ 100, 101, 102, 103, 104 ], "mean_minimum_satisfaction": 0.0522, "mean_sum_satisfaction": 0.9412, "joint_satisfaction_certified": 0, "joint_rational_synthesis_certified": 0, "exact_joint_satisfaction_certified": 0, "mean_training_seconds": 5900.034220169764 }, "territory/ippo/no_cer": { "seeds": [ 100, 101, 102, 103, 104 ], "mean_minimum_satisfaction": 0.359, "mean_sum_satisfaction": 0.9788, "joint_satisfaction_certified": 3, "joint_rational_synthesis_certified": 0, "exact_joint_satisfaction_certified": 3, "mean_training_seconds": 5592.238833738864 }, "territory/ippo/rm_shaping": { "seeds": [ 100, 101, 102, 103, 104 ], "mean_minimum_satisfaction": 0.17880000000000001, "mean_sum_satisfaction": 0.9826, "joint_satisfaction_certified": 1, "joint_rational_synthesis_certified": 0, "exact_joint_satisfaction_certified": 1, "mean_training_seconds": 5780.509105457738 } }, "profiles": [ { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "cer", "seed": 100, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "b62afe5b7c116b79540a59fe9740c7fb8eb4a34e1c902551e5fb234efc8ee269", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2348.343902438879, "eval_seconds": 27.33870017528534, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/cer/seed_100/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "cer", "seed": 101, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "df5de9dcb54d5e5cb7edfacc4a4c248ba629c60570e73e198cd326bf53e7f1ce", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2618.05798882246, "eval_seconds": 27.688456535339355, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/cer/seed_101/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "cer", "seed": 102, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "66fe344226b254df7321ea7555a909fe34d550c1e12db75ebcfcee9c6fa51373", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2212.029641315341, "eval_seconds": 22.298952735960484, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/cer/seed_102/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "cer", "seed": 103, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "cdbc3df86e3347175fda6c28465ce00751c96e710ce765c77d707fa55936be85", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2801.777236253023, "eval_seconds": 29.602483950555325, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/cer/seed_103/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "cer", "seed": 104, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "ad9fb8bb09a57018bd5adafa70cfa232ccee720a0c08eb0b9422c4b7b826e43a", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.999, "successes": 999, "lower": 0.935474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 0.999, "successes": 999, "lower": 0.935474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 0.999, "sum_satisfaction": 1.998, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2885.6580434069037, "eval_seconds": 29.85876737535, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/cer/seed_104/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.01117941664748523, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9888205833525148, "probability_interval": [ 0.9888205833525148, 1.0 ], "gain_upper": 0.01117941664748523, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9888205833525148, "probability_interval": [ 0.9888205833525148, 1.0 ], "gain_upper": 0.01117941664748523, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "corrected", "seed": 100, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "e3374098e6942c775c4677402b388d5c12a84a146f72b9f0188e52fc4739f441", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2395.0265142917633, "eval_seconds": 26.90594481676817, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/corrected/seed_100/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "corrected", "seed": 101, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "8cd332b3c5f2ab64aa63cb2147c399ca477ccaf6d77d984ba6b42c4a0da6c7fb", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2527.739191748202, "eval_seconds": 25.24500449746847, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/corrected/seed_101/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "corrected", "seed": 102, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "84df5dda7c37133a63e2f6a1528d49afa7d0de02d0bfefd3a249afab8e6d31f8", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2246.365330442786, "eval_seconds": 24.368098087608814, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/corrected/seed_102/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "corrected", "seed": 103, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "9b6ce9276522a30d13a5c143fb1b8392b875c17a2d38d1521484d0feef61576a", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2226.0647543370724, "eval_seconds": 26.464444912970066, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/corrected/seed_103/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "corrected", "seed": 104, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "cb240da8f373547a2132a6940e97433104f8b0bcfe9920f6b2e8a90fbdeabc18", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2234.625880777836, "eval_seconds": 23.945904664695263, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/corrected/seed_104/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "critic_only", "seed": 100, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "d8d29ac42b087737c66fa17f8326e80c71002c877e824434e2e3f62a21d7597a", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2676.3229393586516, "eval_seconds": 26.736789748072624, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/critic_only/seed_100/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "critic_only", "seed": 101, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "f63658e5db5c28c6c7030dd9d280af211934c7061ba16671f9d47672a8ff0643", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2596.8990314751863, "eval_seconds": 26.823927521705627, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/critic_only/seed_101/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "critic_only", "seed": 102, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "bc333a7fa00f0ada55cae399d6b1ee4531bbff6e06268fd365cc6564bb95e13a", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2287.1483531221747, "eval_seconds": 24.614175014197826, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/critic_only/seed_102/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "critic_only", "seed": 103, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "d271bb64873d980f3392a7c8970f12ada411d24371f9048f21a0ff25a74780a2", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2774.740619830787, "eval_seconds": 30.638437882065773, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/critic_only/seed_103/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "critic_only", "seed": 104, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "1c558a5dc6c1ef118d89b7cc5425521886ff259fa9847c7cec8d3e4a4b17978c", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 2637.0088985934854, "eval_seconds": 27.751974873244762, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/critic_only/seed_104/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "no_cer", "seed": 100, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "ef03b5ac51edcb14801f7abdfa26a0b312a84840962bb6a69213e703c8315d84", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.024, "successes": 24, "lower": 0.0, "upper": 0.08752521581540601, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.024, "successes": 24, "lower": 0.0, "upper": 0.08752521581540601, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.024, "sum_satisfaction": 0.048, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 1836.63276027143, "eval_seconds": 29.287711679935455, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/no_cer/seed_100/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9897291013031624, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.01027089869683759, "probability_interval": [ 0.01027089869683759, 1.0 ], "gain_upper": 0.9897291013031624, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.01027089869683759, "probability_interval": [ 0.01027089869683759, 1.0 ], "gain_upper": 0.9897291013031624, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "no_cer", "seed": 101, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "ae01df9e8858ae35369b47e8c024fe508198bfe3671a7c710eddbddb783d43de", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 1659.571686834097, "eval_seconds": 28.07329823076725, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/no_cer/seed_101/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "no_cer", "seed": 102, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "b592d1ef9ceef43a51d8afe6d537cbbc51679603097b12f5d64581f9ae29081a", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 1388.9916122704744, "eval_seconds": 23.905347287654877, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/no_cer/seed_102/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "no_cer", "seed": 103, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "91c189c4f01843e1e578634e813b990aec07aa3e59146431eb44b6c7e1f71a9e", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false }, "player_1": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.063525215815406, "threshold": 0.9, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 1671.0564298555255, "eval_seconds": 27.489855244755745, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/no_cer/seed_103/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "no_cer", "seed": 104, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "2a891edbca192787240040cde4452ab055ee337d5024c66e880d2d48b8ce254c", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 1637.9284048229456, "eval_seconds": 28.929836124181747, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/no_cer/seed_104/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "rm_shaping", "seed": 100, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "d3c4533ed6879034eab86daa8f6073526a0837651308b89957ec5e4834deac80", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 1722.5220955237746, "eval_seconds": 29.976837791502476, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/rm_shaping/seed_100/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "rm_shaping", "seed": 101, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "bf3c2e2e19aa407cba343abe4554a7e2f48eef0b75acdbedf469e28e0d990869", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 1619.0521863549948, "eval_seconds": 25.414654225111008, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/rm_shaping/seed_101/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "rm_shaping", "seed": 102, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "cc789a4a4063c6dce8d205f9ded339ef19ad98370f8f0b9e57916f8c65bdfee7", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 1362.0298161879182, "eval_seconds": 23.50325182080269, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/rm_shaping/seed_102/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "rm_shaping", "seed": 103, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "074e00ed8ca4bc754425112c655ae291d73a564eaeb6598355e3d989aa79d445", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 1387.8337820842862, "eval_seconds": 23.586165443062782, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/rm_shaping/seed_103/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "stag_hunt", "learner": "pr2_ippo", "mode": "rm_shaping", "seed": 104, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "4fbca7ca8bb1f5566b09c25b1b46bec39d8d97f1fc2a7c1cd7a7544ac2d66497", "eval_episodes": 1000, "family_size": 40, "per_agent": { "player_0": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true }, "player_1": { "p_hat": 1.0, "successes": 1000, "lower": 0.936474784184594, "upper": 1.0, "threshold": 0.9, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 1.0, "sum_satisfaction": 2.0, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 9606 }, "player_1": { "leaves": 12, "parameters": 9606 } }, "train_seconds": 1384.1231153607368, "eval_seconds": 23.23365420103073, "training_steps": 500224, "gamma": 0.99, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/stag_hunt/pr2_ippo/rm_shaping/seed_104/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.008694288735234146, "epsilon_ne_certified": true, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": true, "per_agent": { "player_0": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true }, "player_1": { "utility_lower": 0.9913057112647659, "probability_interval": [ 0.9913057112647659, 1.0 ], "gain_upper": 0.008694288735234146, "satisfaction_certified": true, "rational_synthesis_certified": true } } } }, { "case": "territory", "learner": "ippo", "mode": "cer", "seed": 100, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "67bac5921c5eb40534304434c29bc495f88e3bdfd3bde3c1333bdf267c3e092c", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.924, "successes": 924, "lower": 0.8623522001222181, "upper": 0.985647799877782, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.064, "successes": 64, "lower": 0.002352200122218144, "upper": 0.12564779987778185, "threshold": 0.2, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.064, "sum_satisfaction": 0.988, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6484.838539218064, "eval_seconds": 1562.658158404287, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/cer/seed_100/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9604007669630575, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.8894785189748322, "probability_interval": [ 0.8894785189748322, 1.0 ], "gain_upper": 0.11052148102516779, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.039599233036942454, "probability_interval": [ 0.039599233036942454, 1.0 ], "gain_upper": 0.9604007669630575, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "cer", "seed": 101, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "a2041a2a673d0ec38650010eace66d19c528cba9f46682a7405e87c4e7290ae1", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.054, "successes": 54, "lower": 0.0, "upper": 0.11564779987778186, "threshold": 0.25, "certified": false }, "player_1": { "p_hat": 0.707, "successes": 707, "lower": 0.6453522001222181, "upper": 0.7686477998777819, "threshold": 0.2, "certified": true } }, "jointly_certified": false, "minimum_satisfaction": 0.054, "sum_satisfaction": 0.761, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6366.937688797712, "eval_seconds": 1565.1352970525622, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/cer/seed_101/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9681990787880288, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.03180092121197123, "probability_interval": [ 0.03180092121197123, 1.0 ], "gain_upper": 0.9681990787880288, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.6530323929752642, "probability_interval": [ 0.6530323929752642, 1.0 ], "gain_upper": 0.34696760702473584, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "cer", "seed": 102, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "be5226cd62fdd6cfee1834b22e8af98a0c1caf539636cba4df0aa4ec78d4cd80", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.945, "successes": 945, "lower": 0.883352200122218, "upper": 1.0, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.044, "successes": 44, "lower": 0.0, "upper": 0.10564779987778186, "threshold": 0.2, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.044, "sum_satisfaction": 0.989, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6394.531883973628, "eval_seconds": 1629.09592346102, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/cer/seed_102/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9757529839191352, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.9144346839092642, "probability_interval": [ 0.9144346839092642, 1.0 ], "gain_upper": 0.08556531609073581, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.02424701608086482, "probability_interval": [ 0.02424701608086482, 1.0 ], "gain_upper": 0.9757529839191352, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "cer", "seed": 103, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "74e2d1ab569a30c0a811358b28da4e3c625af70088f76d4066888d5842dbdd34", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.131, "successes": 131, "lower": 0.06935220012221815, "upper": 0.19264779987778186, "threshold": 0.25, "certified": false }, "player_1": { "p_hat": 0.749, "successes": 749, "lower": 0.6873522001222181, "upper": 0.8106477998777819, "threshold": 0.2, "certified": true } }, "jointly_certified": false, "minimum_satisfaction": 0.131, "sum_satisfaction": 0.88, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 5107.398953057826, "eval_seconds": 1318.2091859653592, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/cer/seed_103/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9044612227706175, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.09553877722938256, "probability_interval": [ 0.09553877722938256, 1.0 ], "gain_upper": 0.9044612227706175, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.6970845290781489, "probability_interval": [ 0.6970845290781489, 1.0 ], "gain_upper": 0.3029154709218511, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "cer", "seed": 104, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "4a875892f5c0984fcb87f19d44b2416bbe2c5547fe69eb7d736c8dd67d98a4fe", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.542, "successes": 542, "lower": 0.4803522001222182, "upper": 0.6036477998777819, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.416, "successes": 416, "lower": 0.35435220012221813, "upper": 0.47764779987778183, "threshold": 0.2, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 0.416, "sum_satisfaction": 0.958, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 5065.679041676223, "eval_seconds": 1300.8876407071948, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/cer/seed_104/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.6396016549564267, "epsilon_ne_certified": false, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.4846573202428179, "probability_interval": [ 0.4846573202428179, 1.0 ], "gain_upper": 0.5153426797571821, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.3603983450435732, "probability_interval": [ 0.3603983450435732, 1.0 ], "gain_upper": 0.6396016549564267, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "corrected", "seed": 100, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "64f460919ff6439b7913bdea3cae9679d2c920ccbb628cff99fd7716000f1cb7", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.229, "successes": 229, "lower": 0.16735220012221816, "upper": 0.2906477998777819, "threshold": 0.25, "certified": false }, "player_1": { "p_hat": 0.733, "successes": 733, "lower": 0.6713522001222181, "upper": 0.7946477998777819, "threshold": 0.2, "certified": true } }, "jointly_certified": false, "minimum_satisfaction": 0.229, "sum_satisfaction": 0.962, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6508.442820666358, "eval_seconds": 1577.5301396562718, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/corrected/seed_100/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.8166982860388764, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.1833017139611236, "probability_interval": [ 0.1833017139611236, 1.0 ], "gain_upper": 0.8166982860388764, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.6802358191887314, "probability_interval": [ 0.6802358191887314, 1.0 ], "gain_upper": 0.31976418081126856, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "corrected", "seed": 101, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "2fb7312282e79ac9c94c0135365d00c02f02e37cbef06289962be403c6581fd9", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.923, "successes": 923, "lower": 0.8613522001222181, "upper": 0.984647799877782, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.031, "successes": 31, "lower": 0.0, "upper": 0.09264779987778185, "threshold": 0.2, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.031, "sum_satisfaction": 0.9540000000000001, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6656.766051992774, "eval_seconds": 1683.6616984456778, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/corrected/seed_101/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9850639828136941, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.8883088453020243, "probability_interval": [ 0.8883088453020243, 1.0 ], "gain_upper": 0.11169115469797575, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.014936017186305882, "probability_interval": [ 0.014936017186305882, 1.0 ], "gain_upper": 0.9850639828136941, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "corrected", "seed": 102, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "cbbcc9fa2bede95c3640c5702aafbcafbd83f62f9a1a86d995718b508b952d18", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.247, "successes": 247, "lower": 0.18535220012221815, "upper": 0.30864779987778185, "threshold": 0.25, "certified": false }, "player_1": { "p_hat": 0.479, "successes": 479, "lower": 0.41735220012221813, "upper": 0.5406477998777819, "threshold": 0.2, "certified": true } }, "jointly_certified": false, "minimum_satisfaction": 0.247, "sum_satisfaction": 0.726, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 8557.540979221463, "eval_seconds": 1937.0845663323998, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/corrected/seed_102/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.8001124402767541, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.1998875597232459, "probability_interval": [ 0.1998875597232459, 1.0 ], "gain_upper": 0.8001124402767541, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.4220764515067317, "probability_interval": [ 0.4220764515067317, 1.0 ], "gain_upper": 0.5779235484932683, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "corrected", "seed": 103, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "68aaaf1a14b2f65e18158ab5bba7431dead60d413685408580c5049209315e95", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.734, "successes": 734, "lower": 0.6723522001222181, "upper": 0.7956477998777819, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.171, "successes": 171, "lower": 0.10935220012221816, "upper": 0.23264779987778186, "threshold": 0.2, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.171, "sum_satisfaction": 0.905, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 5111.825251638889, "eval_seconds": 1324.7467078790069, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/corrected/seed_103/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.8692397048726761, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.6812863734259216, "probability_interval": [ 0.6812863734259216, 1.0 ], "gain_upper": 0.3187136265740784, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.1307602951273239, "probability_interval": [ 0.1307602951273239, 1.0 ], "gain_upper": 0.8692397048726761, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "corrected", "seed": 104, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "4fd9cc6607011365cf3247086cd51a65bbcbe134929fe28fbbd2108c3d6edab8", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.991, "successes": 991, "lower": 0.9293522001222181, "upper": 1.0, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.008, "successes": 8, "lower": 0.0, "upper": 0.06964779987778186, "threshold": 0.2, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.008, "sum_satisfaction": 0.999, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 5017.611191160977, "eval_seconds": 1288.6375302225351, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/corrected/seed_104/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9985070480853762, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.9747303430359303, "probability_interval": [ 0.9747303430359303, 1.0 ], "gain_upper": 0.025269656964069687, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.0014929519146238154, "probability_interval": [ 0.0014929519146238154, 1.0 ], "gain_upper": 0.9985070480853762, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "critic_only", "seed": 100, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "76a4b52de70af1890d6774661c60c137647e5886e371e36ad9a79ba870454d71", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.0, "successes": 0, "lower": 0.0, "upper": 0.06164779987778186, "threshold": 0.25, "certified": false }, "player_1": { "p_hat": 0.999, "successes": 999, "lower": 0.9373522001222181, "upper": 1.0, "threshold": 0.2, "certified": true } }, "jointly_certified": false, "minimum_satisfaction": 0.0, "sum_satisfaction": 0.999, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6465.878381901886, "eval_seconds": 1556.0090998131782, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/critic_only/seed_100/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 1.0, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.0, "probability_interval": [ 0.0, 1.0 ], "gain_upper": 1.0, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.9888205833525148, "probability_interval": [ 0.9888205833525148, 1.0 ], "gain_upper": 0.01117941664748523, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "critic_only", "seed": 101, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "53e6accc86fd69798f46fa724e15e7257d6d34c57e7408b73c0390b3c1c87909", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.22, "successes": 220, "lower": 0.15835220012221815, "upper": 0.2816477998777819, "threshold": 0.25, "certified": false }, "player_1": { "p_hat": 0.54, "successes": 540, "lower": 0.4783522001222182, "upper": 0.6016477998777819, "threshold": 0.2, "certified": true } }, "jointly_certified": false, "minimum_satisfaction": 0.22, "sum_satisfaction": 0.76, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6402.205967985094, "eval_seconds": 1645.9447301477194, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/critic_only/seed_101/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.8249455552880494, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.17505444471195059, "probability_interval": [ 0.17505444471195059, 1.0 ], "gain_upper": 0.8249455552880494, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.4826567946069178, "probability_interval": [ 0.4826567946069178, 1.0 ], "gain_upper": 0.5173432053930822, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "critic_only", "seed": 102, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "41378547d7582f85161bf65977d266c45b3f4f370e6e3f02da572a12a81e908a", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.968, "successes": 968, "lower": 0.9063522001222181, "upper": 1.0, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.02, "successes": 20, "lower": 0.0, "upper": 0.08164779987778185, "threshold": 0.2, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.02, "sum_satisfaction": 0.988, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6701.542805373669, "eval_seconds": 1721.1085481550545, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/critic_only/seed_102/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9922346329925231, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.9430199200852487, "probability_interval": [ 0.9430199200852487, 1.0 ], "gain_upper": 0.05698007991475129, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.00776536700747689, "probability_interval": [ 0.00776536700747689, 1.0 ], "gain_upper": 0.9922346329925231, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "critic_only", "seed": 103, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "bba2f94191c03d67857f1abd0542fea7b79023a65ce90437157b80c6cd45090a", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.963, "successes": 963, "lower": 0.9013522001222181, "upper": 1.0, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.018, "successes": 18, "lower": 0.0, "upper": 0.07964779987778185, "threshold": 0.2, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.018, "sum_satisfaction": 0.981, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 5037.8036170601845, "eval_seconds": 1314.680282726884, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/critic_only/seed_103/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9934306596958314, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.9366470787280329, "probability_interval": [ 0.9366470787280329, 1.0 ], "gain_upper": 0.06335292127196712, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.006569340304168525, "probability_interval": [ 0.006569340304168525, 1.0 ], "gain_upper": 0.9934306596958314, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "critic_only", "seed": 104, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "19323e467650596c5001998922b442dbf401e1c9af7b8a483f6b51f4e07fdf17", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.003, "successes": 3, "lower": 0.0, "upper": 0.06464779987778185, "threshold": 0.25, "certified": false }, "player_1": { "p_hat": 0.975, "successes": 975, "lower": 0.9133522001222181, "upper": 1.0, "threshold": 0.2, "certified": true } }, "jointly_certified": false, "minimum_satisfaction": 0.003, "sum_satisfaction": 0.978, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 4892.740328527987, "eval_seconds": 1184.7109229117632, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/critic_only/seed_104/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9998984564875669, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.00010154351243304963, "probability_interval": [ 0.00010154351243304963, 1.0 ], "gain_upper": 0.9998984564875669, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.9521572872703304, "probability_interval": [ 0.9521572872703304, 1.0 ], "gain_upper": 0.047842712729669556, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "no_cer", "seed": 100, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "c49d60c44c0c4325a01b1ba04e3259f6c605f006ff6ced6ca2c00f805d9b66cf", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.807, "successes": 807, "lower": 0.7453522001222181, "upper": 0.868647799877782, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.178, "successes": 178, "lower": 0.11635220012221814, "upper": 0.23964779987778184, "threshold": 0.2, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.178, "sum_satisfaction": 0.9850000000000001, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6974.16438806802, "eval_seconds": 1755.403563410975, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/no_cer/seed_100/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.8629811416010253, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.7589511744421104, "probability_interval": [ 0.7589511744421104, 1.0 ], "gain_upper": 0.24104882555788965, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.1370188583989747, "probability_interval": [ 0.1370188583989747, 1.0 ], "gain_upper": 0.8629811416010253, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "no_cer", "seed": 101, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "f8102db830461df35d69ed769441bbb9def8fff5530e7fd3c263dcf7624c1202", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.274, "successes": 274, "lower": 0.21235220012221817, "upper": 0.33564779987778187, "threshold": 0.25, "certified": false }, "player_1": { "p_hat": 0.709, "successes": 709, "lower": 0.6473522001222181, "upper": 0.7706477998777819, "threshold": 0.2, "certified": true } }, "jointly_certified": false, "minimum_satisfaction": 0.274, "sum_satisfaction": 0.983, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6031.014965467155, "eval_seconds": 1628.2661053091288, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/no_cer/seed_101/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.7750219722884153, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.2249780277115847, "probability_interval": [ 0.2249780277115847, 1.0 ], "gain_upper": 0.7750219722884153, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.6551175797895765, "probability_interval": [ 0.6551175797895765, 1.0 ], "gain_upper": 0.3448824202104235, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "no_cer", "seed": 102, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "3d8936993ddb6c0e7b6f5dc16f481339ab98ece3feea8ad18ffb771e8a0b0b0c", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.522, "successes": 522, "lower": 0.46035220012221817, "upper": 0.5836477998777819, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.449, "successes": 449, "lower": 0.38735220012221816, "upper": 0.5106477998777819, "threshold": 0.2, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 0.449, "sum_satisfaction": 0.9710000000000001, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 5968.610487557948, "eval_seconds": 1812.7270233929157, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/no_cer/seed_102/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.6074078898586854, "epsilon_ne_certified": false, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.46469302040504384, "probability_interval": [ 0.46469302040504384, 1.0 ], "gain_upper": 0.5353069795949561, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.3925921101413146, "probability_interval": [ 0.3925921101413146, 1.0 ], "gain_upper": 0.6074078898586854, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "no_cer", "seed": 103, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "1f383caca6e52b637cf02e16c3409ff1d65069ea4b2f32d0aa98a15a00d8491b", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.454, "successes": 454, "lower": 0.39235220012221816, "upper": 0.5156477998777819, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.532, "successes": 532, "lower": 0.4703522001222182, "upper": 0.5936477998777819, "threshold": 0.2, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 0.454, "sum_satisfaction": 0.986, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 4368.338847950101, "eval_seconds": 1265.9411757290363, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/no_cer/seed_103/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.6025080912726462, "epsilon_ne_certified": false, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.3974919087273539, "probability_interval": [ 0.3974919087273539, 1.0 ], "gain_upper": 0.6025080912726462, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.4746638057407484, "probability_interval": [ 0.4746638057407484, 1.0 ], "gain_upper": 0.5253361942592516, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "no_cer", "seed": 104, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "699693db29b0ddea8ec9ca44df90913e6cfa9d07c10a2291cb992da6c8e6ef57", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.529, "successes": 529, "lower": 0.4673522001222182, "upper": 0.5906477998777819, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.44, "successes": 440, "lower": 0.37835220012221815, "upper": 0.5016477998777819, "threshold": 0.2, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 0.44, "sum_satisfaction": 0.9690000000000001, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 4619.0654796510935, "eval_seconds": 1218.6861491799355, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/no_cer/seed_104/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.616213053313757, "epsilon_ne_certified": false, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.47167018740143185, "probability_interval": [ 0.47167018740143185, 1.0 ], "gain_upper": 0.5283298125985681, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.383786946686243, "probability_interval": [ 0.383786946686243, 1.0 ], "gain_upper": 0.616213053313757, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "rm_shaping", "seed": 100, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "ec17be375f0a5c6f8d2be94cce449654166bdf2ea253d13cb2fe02f0f74eea85", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.59, "successes": 590, "lower": 0.5283522001222181, "upper": 0.6516477998777819, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.366, "successes": 366, "lower": 0.30435220012221814, "upper": 0.42764779987778184, "threshold": 0.2, "certified": true } }, "jointly_certified": true, "minimum_satisfaction": 0.366, "sum_satisfaction": 0.956, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6312.894241899252, "eval_seconds": 1738.4824618026614, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/rm_shaping/seed_100/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.6878865425907175, "epsilon_ne_certified": false, "joint_satisfaction_certified": true, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.5329466079878828, "probability_interval": [ 0.5329466079878828, 1.0 ], "gain_upper": 0.46705339201211715, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.3121134574092826, "probability_interval": [ 0.3121134574092826, 1.0 ], "gain_upper": 0.6878865425907175, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "rm_shaping", "seed": 101, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "ace96971fec458ad1f13241f74ac2a7f862c33fab3a942f2086c7d6bab21a484", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.873, "successes": 873, "lower": 0.8113522001222181, "upper": 0.9346477998777819, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.118, "successes": 118, "lower": 0.056352200122218137, "upper": 0.17964779987778184, "threshold": 0.2, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.118, "sum_satisfaction": 0.991, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 5906.288038849831, "eval_seconds": 1432.9165427982807, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/rm_shaping/seed_101/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9156734643469071, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.8313018047780754, "probability_interval": [ 0.8313018047780754, 1.0 ], "gain_upper": 0.16869819522192464, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.08432653565309285, "probability_interval": [ 0.08432653565309285, 1.0 ], "gain_upper": 0.9156734643469071, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "rm_shaping", "seed": 102, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "17f68cd461acbb61b8181572049f632e0dd59844a9851683d8a0a6e4b01e14a7", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.817, "successes": 817, "lower": 0.755352200122218, "upper": 0.8786477998777819, "threshold": 0.25, "certified": true }, "player_1": { "p_hat": 0.169, "successes": 169, "lower": 0.10735220012221816, "upper": 0.23064779987778186, "threshold": 0.2, "certified": false } }, "jointly_certified": false, "minimum_satisfaction": 0.169, "sum_satisfaction": 0.986, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 5817.911510653794, "eval_seconds": 1609.686750434339, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/rm_shaping/seed_102/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.8710231617782611, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.7697609805019471, "probability_interval": [ 0.7697609805019471, 1.0 ], "gain_upper": 0.23023901949805292, "satisfaction_certified": true, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.12897683822173883, "probability_interval": [ 0.12897683822173883, 1.0 ], "gain_upper": 0.8710231617782611, "satisfaction_certified": false, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "rm_shaping", "seed": 103, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "919e74d27757e3afd0dac81e5e76125545a5e842a3539f971e15dc29c22c94d9", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.016, "successes": 16, "lower": 0.0, "upper": 0.07764779987778186, "threshold": 0.25, "certified": false }, "player_1": { "p_hat": 0.979, "successes": 979, "lower": 0.9173522001222181, "upper": 1.0, "threshold": 0.2, "certified": true } }, "jointly_certified": false, "minimum_satisfaction": 0.016, "sum_satisfaction": 0.995, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 4811.639261499047, "eval_seconds": 1322.9819707721472, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/rm_shaping/seed_103/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.9945808803160312, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.005419119683968836, "probability_interval": [ 0.005419119683968836, 1.0 ], "gain_upper": 0.9945808803160312, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.9575254248480523, "probability_interval": [ 0.9575254248480523, 1.0 ], "gain_upper": 0.04247457515194775, "satisfaction_certified": true, "rational_synthesis_certified": false } } } }, { "case": "territory", "learner": "ippo", "mode": "rm_shaping", "seed": 104, "source_commit": "612145f915d524d75c600f14db763544272e1d83", "source_hash": "ea456fe75f0fa42731f35a6e68197795a5cb424d4f2e3a24f6dfa8ac08039501", "configuration_fingerprint": "9ed3fede36ce1f7ba4c8000f537466612793e09b319b9bed092359d4e96213c9", "eval_episodes": 1000, "family_size": 25, "per_agent": { "player_0": { "p_hat": 0.225, "successes": 225, "lower": 0.16335220012221816, "upper": 0.2866477998777819, "threshold": 0.25, "certified": false }, "player_1": { "p_hat": 0.76, "successes": 760, "lower": 0.6983522001222181, "upper": 0.8216477998777819, "threshold": 0.2, "certified": true } }, "jointly_certified": false, "minimum_satisfaction": 0.225, "sum_satisfaction": 0.985, "parameter_validation": { "player_0": { "leaves": 12, "parameters": 517386 }, "player_1": { "leaves": 12, "parameters": 517386 } }, "train_seconds": 6053.812474386767, "eval_seconds": 1653.3267205692828, "training_steps": 200192, "gamma": 1.0, "artifact_path": "/vol/gpudata/oja24-pltlf/exports/optimisations/cer-study-20261012/full/territory/ippo/rm_shaping/seed_104/result.json", "payoff_ceiling_certificate_family155": { "method": "payoff_ceiling_clopper_pearson", "utility": "p_for_lower_bounds_1_minus_p_for_upper_bounds", "scope": "initial_state_epsilon_ne", "epsilon": 0.05, "delta": 0.05, "confidence": 0.95, "family_size": 155, "num_comparisons": 310, "per_agent_delta": 0.00016129032258064516, "exploitability_upper": 0.8203676232209696, "epsilon_ne_certified": false, "joint_satisfaction_certified": false, "joint_rational_synthesis_certified": false, "per_agent": { "player_0": { "utility_lower": 0.1796323767790305, "probability_interval": [ 0.1796323767790305, 1.0 ], "gain_upper": 0.8203676232209696, "satisfaction_certified": false, "rational_synthesis_certified": false }, "player_1": { "utility_lower": 0.7087189875293642, "probability_interval": [ 0.7087189875293642, 1.0 ], "gain_upper": 0.29128101247063576, "satisfaction_certified": true, "rational_synthesis_certified": false } } } } ] }