{
  "seed": 42,
  "scope": "candidate-policy and deterministic agent mini-suite mechanism validation",
  "post_training": {
    "opsd": {
      "dataset": "gsm8k-candidate",
      "examples": 128,
      "steps": 120,
      "baseline_accuracy": 0.25,
      "final_accuracy": 0.90625,
      "mean_reward": 0.9009100667879821,
      "online_teacher_calls": 480
    },
    "opcd": {
      "dataset": "gsm8k-candidate",
      "examples": 128,
      "steps": 120,
      "baseline_accuracy": 0.25,
      "final_accuracy": 0.96875,
      "mean_reward": 0.9509380677139144,
      "online_teacher_calls": 480
    }
  },
  "agentic_rl": {
    "loop": {
      "benchmark": "planbench-mini",
      "episodes": 120,
      "joint_success": 1.0,
      "average_cost": 2.2,
      "off_policy_reuses": 476,
      "leave_one_out_updates": 480,
      "per_token_clips": 120
    },
    "webagent-r1": {
      "benchmark": "scalemcp-mini",
      "episodes": 120,
      "joint_success": 1.0,
      "average_cost": 2.6,
      "context_tokens_compressed": 360,
      "parallel_trajectory_groups": 120,
      "multi_turn_group_updates": 120
    },
    "mua-rl": {
      "benchmark": "scalemcp-mini",
      "episodes": 120,
      "joint_success": 1.0,
      "average_cost": 2.25,
      "simulated_user_turns": 360,
      "intent_refinements": 240,
      "real_tool_responses": 360,
      "task_completion_rewards": 120
    }
  },
  "schema_version": 2,
  "manifest_ref": "experiments:omitted-agentic-rl-opd-seed42",
  "evaluation_protocol": {
    "tier": "l1_mechanism",
    "seeds": [
      42
    ],
    "formal_comparison": false,
    "claim_policy": "single/few-seed smoke result; do not claim a stable improvement"
  },
  "provenance": {
    "artifact_path": "docs/experiments/omitted-agentic-rl-opd-seed42.json",
    "historical_migration": "historical-metrics-v2-2026-08-09",
    "original_code_commit": "not recorded",
    "dataset_fingerprint": "not recorded in historical artifact"
  }
}