datasets/fixed_levels/qwen_swarm_eval_fixed_levels.json

{
  "avg_compactness_reward": 0.0,
  "avg_connectivity_gain_reward": 0.16666666666666666,
  "avg_connectivity_reward": 0.16999999999999998,
  "avg_diversity_reward": 0.1157777777777778,
  "avg_entity_informativeness_reward": -0.02824631570420193,
  "avg_format_reward": 0.14999999999999997,
  "avg_graph_f1": 0.8492063492063492,
  "avg_knowledge_carrier_reward": 0.5,
  "avg_knowledge_indexing_reward": 0.07400000000000001,
  "avg_relation_informativeness_reward": 0.06905976285357758,
  "avg_reward": 4.285384567790942,
  "avg_soft_shaping_reward": 0.24999999999999994,
  "avg_spawn_count": 4.0,
  "avg_spawn_critical_steps": 6.0,
  "avg_steps_to_solution": 9.0,
  "deanonymization_accuracy": 1.0,
  "leaderboard_score": 0.8565775118852701,
  "retrieval_signal": 0.7009000000000001,
  "spawn_completion_rate": 1.0,
  "spawn_signal": 0.6666666666666666,
  "structural_signal": 0.5846960227632085,
  "task_success_rate": 1.0,
  "tool_efficiency": 0.5
}