{
  "arms": {
    "base": {
      "all": {
        "abstention_rate": 0.0,
        "answer_tokens": 18125,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": 0.0,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.003472222222222222,
        "injected_tokens": 576,
        "logical_model_input_tokens": 534768,
        "macro_use_rate_correct_candidates": 0.0,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 4.994318181818182,
        "mean_sampled_tokens_per_completion": 255.54861111111111,
        "mean_sampled_tokens_per_task": 6133.166666666667,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 4.994318181818182,
        "n_samples": 288,
        "n_tasks": 12,
        "oracle_hidden_coverage": 0.08333333333333333,
        "parse_rate": 0.6111111111111112,
        "sampled_plus_unique_prompt_tokens": 83575,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.08333333333333333,
        "selected_visible_pass_rate": 0.08333333333333333,
        "thinking_tokens": 55296,
        "total_sampled_tokens": 73598,
        "total_unique_prompt_tokens": 9977,
        "truncation_rate": 0.3854166666666667,
        "valid_program_rate": 0.6111111111111112,
        "visible_pass_rate": 0.003472222222222222
      },
      "smoke_no_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 9422,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": 0.0,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.006944444444444444,
        "injected_tokens": 288,
        "logical_model_input_tokens": 267696,
        "macro_use_rate_correct_candidates": 0.0,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 258.0208333333333,
        "mean_sampled_tokens_per_task": 6192.5,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 144,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.16666666666666666,
        "parse_rate": 0.5902777777777778,
        "sampled_plus_unique_prompt_tokens": 42150,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.16666666666666666,
        "selected_visible_pass_rate": 0.16666666666666666,
        "thinking_tokens": 27648,
        "total_sampled_tokens": 37155,
        "total_unique_prompt_tokens": 4995,
        "truncation_rate": 0.4097222222222222,
        "valid_program_rate": 0.5902777777777778,
        "visible_pass_rate": 0.006944444444444444
      },
      "smoke_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 8703,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 288,
        "logical_model_input_tokens": 267072,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 4.989010989010989,
        "mean_sampled_tokens_per_completion": 253.07638888888889,
        "mean_sampled_tokens_per_task": 6073.833333333333,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 4.989010989010989,
        "n_samples": 144,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.6319444444444444,
        "sampled_plus_unique_prompt_tokens": 41425,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 27648,
        "total_sampled_tokens": 36443,
        "total_unique_prompt_tokens": 4982,
        "truncation_rate": 0.3611111111111111,
        "valid_program_rate": 0.6319444444444444,
        "visible_pass_rate": 0.0
      }
    },
    "designed_ceiling": {
      "all": {
        "abstention_rate": 0.0,
        "answer_tokens": 9653,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 288,
        "logical_model_input_tokens": 283800,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 4.916666666666667,
        "mean_expanded_depth_valid": 4.975,
        "mean_sampled_tokens_per_completion": 259.6041666666667,
        "mean_sampled_tokens_per_task": 3115.25,
        "mean_surface_depth_selected": 4.916666666666667,
        "mean_surface_depth_valid": 4.975,
        "n_samples": 144,
        "n_tasks": 12,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5694444444444444,
        "sampled_plus_unique_prompt_tokens": 48044,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 27648,
        "total_sampled_tokens": 37383,
        "total_unique_prompt_tokens": 10661,
        "truncation_rate": 0.4305555555555556,
        "valid_program_rate": 0.5555555555555556,
        "visible_pass_rate": 0.0
      },
      "smoke_no_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 4873,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142056,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 4.833333333333333,
        "mean_expanded_depth_valid": 4.975609756097561,
        "mean_sampled_tokens_per_completion": 260.25,
        "mean_sampled_tokens_per_task": 3123.0,
        "mean_surface_depth_selected": 4.833333333333333,
        "mean_surface_depth_valid": 4.975609756097561,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5694444444444444,
        "sampled_plus_unique_prompt_tokens": 24075,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18738,
        "total_unique_prompt_tokens": 5337,
        "truncation_rate": 0.4305555555555556,
        "valid_program_rate": 0.5694444444444444,
        "visible_pass_rate": 0.0
      },
      "smoke_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 4780,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 141744,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 4.9743589743589745,
        "mean_sampled_tokens_per_completion": 258.9583333333333,
        "mean_sampled_tokens_per_task": 3107.5,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 4.9743589743589745,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5694444444444444,
        "sampled_plus_unique_prompt_tokens": 23969,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18645,
        "total_unique_prompt_tokens": 5324,
        "truncation_rate": 0.4305555555555556,
        "valid_program_rate": 0.5416666666666666,
        "visible_pass_rate": 0.0
      }
    },
    "mined": {
      "all": {
        "abstention_rate": 0.0,
        "answer_tokens": 9639,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284088,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 4.916666666666667,
        "mean_expanded_depth_valid": 4.987654320987654,
        "mean_sampled_tokens_per_completion": 259.50694444444446,
        "mean_sampled_tokens_per_task": 3114.0833333333335,
        "mean_surface_depth_selected": 4.916666666666667,
        "mean_surface_depth_valid": 4.987654320987654,
        "n_samples": 144,
        "n_tasks": 12,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5694444444444444,
        "sampled_plus_unique_prompt_tokens": 48042,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 27648,
        "total_sampled_tokens": 37369,
        "total_unique_prompt_tokens": 10673,
        "truncation_rate": 0.4305555555555556,
        "valid_program_rate": 0.5625,
        "visible_pass_rate": 0.0
      },
      "smoke_no_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 5097,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142200,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 4.833333333333333,
        "mean_expanded_depth_valid": 4.973684210526316,
        "mean_sampled_tokens_per_completion": 263.31944444444446,
        "mean_sampled_tokens_per_task": 3159.8333333333335,
        "mean_surface_depth_selected": 4.833333333333333,
        "mean_surface_depth_valid": 4.973684210526316,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5277777777777778,
        "sampled_plus_unique_prompt_tokens": 24302,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18959,
        "total_unique_prompt_tokens": 5343,
        "truncation_rate": 0.4722222222222222,
        "valid_program_rate": 0.5277777777777778,
        "visible_pass_rate": 0.0
      },
      "smoke_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 4542,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 141888,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 255.69444444444446,
        "mean_sampled_tokens_per_task": 3068.3333333333335,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.6111111111111112,
        "sampled_plus_unique_prompt_tokens": 23740,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18410,
        "total_unique_prompt_tokens": 5330,
        "truncation_rate": 0.3888888888888889,
        "valid_program_rate": 0.5972222222222222,
        "visible_pass_rate": 0.0
      }
    },
    "mined_hint": {
      "all": {
        "abstention_rate": 0.0,
        "answer_tokens": 10314,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 288,
        "logical_model_input_tokens": 285528,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 264.15972222222223,
        "mean_sampled_tokens_per_task": 3169.9166666666665,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 144,
        "n_tasks": 12,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5277777777777778,
        "sampled_plus_unique_prompt_tokens": 48772,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 27648,
        "total_sampled_tokens": 38039,
        "total_unique_prompt_tokens": 10733,
        "truncation_rate": 0.4652777777777778,
        "valid_program_rate": 0.5277777777777778,
        "visible_pass_rate": 0.0
      },
      "smoke_no_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 5000,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142920,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 262.0,
        "mean_sampled_tokens_per_task": 3144.0,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5555555555555556,
        "sampled_plus_unique_prompt_tokens": 24237,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18864,
        "total_unique_prompt_tokens": 5373,
        "truncation_rate": 0.4444444444444444,
        "valid_program_rate": 0.5555555555555556,
        "visible_pass_rate": 0.0
      },
      "smoke_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 5314,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142608,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 266.31944444444446,
        "mean_sampled_tokens_per_task": 3195.8333333333335,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5,
        "sampled_plus_unique_prompt_tokens": 24535,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 19175,
        "total_unique_prompt_tokens": 5360,
        "truncation_rate": 0.4861111111111111,
        "valid_program_rate": 0.5,
        "visible_pass_rate": 0.0
      }
    },
    "random_0": {
      "all": {
        "abstention_rate": 0.0,
        "answer_tokens": 10097,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284376,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 262.65972222222223,
        "mean_sampled_tokens_per_task": 3151.9166666666665,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 144,
        "n_tasks": 12,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5347222222222222,
        "sampled_plus_unique_prompt_tokens": 48508,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 27648,
        "total_sampled_tokens": 37823,
        "total_unique_prompt_tokens": 10685,
        "truncation_rate": 0.4583333333333333,
        "valid_program_rate": 0.5347222222222222,
        "visible_pass_rate": 0.0
      },
      "smoke_no_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 4859,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142344,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 260.05555555555554,
        "mean_sampled_tokens_per_task": 3120.6666666666665,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5555555555555556,
        "sampled_plus_unique_prompt_tokens": 24073,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18724,
        "total_unique_prompt_tokens": 5349,
        "truncation_rate": 0.4305555555555556,
        "valid_program_rate": 0.5555555555555556,
        "visible_pass_rate": 0.0
      },
      "smoke_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 5238,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142032,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 265.2638888888889,
        "mean_sampled_tokens_per_task": 3183.1666666666665,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5138888888888888,
        "sampled_plus_unique_prompt_tokens": 24435,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 19099,
        "total_unique_prompt_tokens": 5336,
        "truncation_rate": 0.4861111111111111,
        "valid_program_rate": 0.5138888888888888,
        "visible_pass_rate": 0.0
      }
    },
    "random_1": {
      "all": {
        "abstention_rate": 0.0,
        "answer_tokens": 9196,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284088,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 256.46527777777777,
        "mean_sampled_tokens_per_task": 3077.5833333333335,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 144,
        "n_tasks": 12,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.6041666666666666,
        "sampled_plus_unique_prompt_tokens": 47604,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 27648,
        "total_sampled_tokens": 36931,
        "total_unique_prompt_tokens": 10673,
        "truncation_rate": 0.3958333333333333,
        "valid_program_rate": 0.6041666666666666,
        "visible_pass_rate": 0.0
      },
      "smoke_no_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 4983,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142200,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 261.7638888888889,
        "mean_sampled_tokens_per_task": 3141.1666666666665,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5555555555555556,
        "sampled_plus_unique_prompt_tokens": 24190,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18847,
        "total_unique_prompt_tokens": 5343,
        "truncation_rate": 0.4444444444444444,
        "valid_program_rate": 0.5555555555555556,
        "visible_pass_rate": 0.0
      },
      "smoke_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 4213,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 141888,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 251.16666666666666,
        "mean_sampled_tokens_per_task": 3014.0,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.6527777777777778,
        "sampled_plus_unique_prompt_tokens": 23414,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18084,
        "total_unique_prompt_tokens": 5330,
        "truncation_rate": 0.3472222222222222,
        "valid_program_rate": 0.6527777777777778,
        "visible_pass_rate": 0.0
      }
    },
    "random_2": {
      "all": {
        "abstention_rate": 0.0,
        "answer_tokens": 9349,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284088,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 4.988095238095238,
        "mean_sampled_tokens_per_completion": 257.5138888888889,
        "mean_sampled_tokens_per_task": 3090.1666666666665,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 4.988095238095238,
        "n_samples": 144,
        "n_tasks": 12,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5902777777777778,
        "sampled_plus_unique_prompt_tokens": 47755,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 27648,
        "total_sampled_tokens": 37082,
        "total_unique_prompt_tokens": 10673,
        "truncation_rate": 0.4097222222222222,
        "valid_program_rate": 0.5833333333333334,
        "visible_pass_rate": 0.0
      },
      "smoke_no_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 4697,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142200,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 4.975609756097561,
        "mean_sampled_tokens_per_completion": 257.81944444444446,
        "mean_sampled_tokens_per_task": 3093.8333333333335,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 4.975609756097561,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5833333333333334,
        "sampled_plus_unique_prompt_tokens": 23906,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18563,
        "total_unique_prompt_tokens": 5343,
        "truncation_rate": 0.4166666666666667,
        "valid_program_rate": 0.5694444444444444,
        "visible_pass_rate": 0.0
      },
      "smoke_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 4652,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 141888,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 257.2083333333333,
        "mean_sampled_tokens_per_task": 3086.5,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5972222222222222,
        "sampled_plus_unique_prompt_tokens": 23849,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18519,
        "total_unique_prompt_tokens": 5330,
        "truncation_rate": 0.4027777777777778,
        "valid_program_rate": 0.5972222222222222,
        "visible_pass_rate": 0.0
      }
    },
    "random_3": {
      "all": {
        "abstention_rate": 0.0,
        "answer_tokens": 9946,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284088,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 4.9875,
        "mean_sampled_tokens_per_completion": 261.63194444444446,
        "mean_sampled_tokens_per_task": 3139.5833333333335,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 4.9875,
        "n_samples": 144,
        "n_tasks": 12,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5555555555555556,
        "sampled_plus_unique_prompt_tokens": 48348,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 27648,
        "total_sampled_tokens": 37675,
        "total_unique_prompt_tokens": 10673,
        "truncation_rate": 0.4375,
        "valid_program_rate": 0.5555555555555556,
        "visible_pass_rate": 0.0
      },
      "smoke_no_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 4872,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142200,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 4.975609756097561,
        "mean_sampled_tokens_per_completion": 260.25,
        "mean_sampled_tokens_per_task": 3123.0,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 4.975609756097561,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5694444444444444,
        "sampled_plus_unique_prompt_tokens": 24081,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18738,
        "total_unique_prompt_tokens": 5343,
        "truncation_rate": 0.4166666666666667,
        "valid_program_rate": 0.5694444444444444,
        "visible_pass_rate": 0.0
      },
      "smoke_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 5074,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 141888,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 263.0138888888889,
        "mean_sampled_tokens_per_task": 3156.1666666666665,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5416666666666666,
        "sampled_plus_unique_prompt_tokens": 24267,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18937,
        "total_unique_prompt_tokens": 5330,
        "truncation_rate": 0.4583333333333333,
        "valid_program_rate": 0.5416666666666666,
        "visible_pass_rate": 0.0
      }
    },
    "random_4": {
      "all": {
        "abstention_rate": 0.0,
        "answer_tokens": 9591,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284376,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 259.1875,
        "mean_sampled_tokens_per_task": 3110.25,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 144,
        "n_tasks": 12,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5833333333333334,
        "sampled_plus_unique_prompt_tokens": 48008,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 27648,
        "total_sampled_tokens": 37323,
        "total_unique_prompt_tokens": 10685,
        "truncation_rate": 0.4166666666666667,
        "valid_program_rate": 0.5833333333333334,
        "visible_pass_rate": 0.0
      },
      "smoke_no_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 4310,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142344,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 252.5,
        "mean_sampled_tokens_per_task": 3030.0,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.6388888888888888,
        "sampled_plus_unique_prompt_tokens": 23529,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 18180,
        "total_unique_prompt_tokens": 5349,
        "truncation_rate": 0.3611111111111111,
        "valid_program_rate": 0.6388888888888888,
        "visible_pass_rate": 0.0
      },
      "smoke_reuse": {
        "abstention_rate": 0.0,
        "answer_tokens": 5281,
        "false_visible_pass_count": 0,
        "false_visible_pass_rate_all_samples": 0.0,
        "false_visible_pass_rate_given_visible_pass": null,
        "forced_close_rate": 1.0,
        "hidden_pass_sample_rate": 0.0,
        "injected_tokens": 144,
        "logical_model_input_tokens": 142032,
        "macro_use_rate_correct_candidates": null,
        "macro_use_rate_selected": 0.0,
        "macro_use_rate_valid_candidates": 0.0,
        "mean_expanded_depth_selected": 5.0,
        "mean_expanded_depth_valid": 5.0,
        "mean_sampled_tokens_per_completion": 265.875,
        "mean_sampled_tokens_per_task": 3190.5,
        "mean_surface_depth_selected": 5.0,
        "mean_surface_depth_valid": 5.0,
        "n_samples": 72,
        "n_tasks": 6,
        "oracle_hidden_coverage": 0.0,
        "parse_rate": 0.5277777777777778,
        "sampled_plus_unique_prompt_tokens": 24479,
        "selected_false_visible_pass_rate": 0.0,
        "selected_hidden_accuracy": 0.0,
        "selected_visible_pass_rate": 0.0,
        "thinking_tokens": 13824,
        "total_sampled_tokens": 19143,
        "total_unique_prompt_tokens": 5336,
        "truncation_rate": 0.4722222222222222,
        "valid_program_rate": 0.5277777777777778,
        "visible_pass_rate": 0.0
      }
    }
  },
  "bootstrap_repetitions": 10000,
  "bootstrap_seed": 8102,
  "run": "smoke",
  "runner_metadata": {
    "base": {
      "counts": {
        "completions": 288,
        "injected_tokens": 576,
        "logical_model_input_tokens": 534768,
        "requests": 12,
        "sampled_tokens": 73598,
        "stage1_logical_prompt_tokens": 239448,
        "stage2_logical_prompt_tokens": 295320,
        "unique_input_prompt_tokens": 9977
      },
      "engine": {
        "adapter": null,
        "enable_prefix_caching": false,
        "enforce_eager": false,
        "gpu_memory_utilization": 0.9,
        "max_model_len": 16384,
        "max_num_batched_tokens": 32768,
        "max_num_seqs": 128
      },
      "model": "Qwen/Qwen3.5-4B",
      "model_revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a",
      "runner_sha256": "022f956ca5185fb67d5913b21465d1e465f6a1fee09067846731c87718d4527c",
      "sampling": {
        "allow_custom_prompts": false,
        "answer_max_tokens": 128,
        "frequency_penalty": 0.0,
        "greedy": false,
        "logprob_token_ids": [],
        "logprobs": null,
        "max_tokens": 128,
        "min_p": 0.0,
        "n": 24,
        "presence_penalty": 0.0,
        "prompt_logprobs": null,
        "repetition_penalty": 1.0,
        "run_seed": 2701,
        "shuffle_thinking": false,
        "temperature": 0.6,
        "thinking": "budget",
        "thinking_budget": 192,
        "top_k": 20,
        "top_p": 0.95
      },
      "schema_version": 2,
      "timing": {
        "generation_seconds": 67.48177834041417,
        "model_load_seconds": 17.697869495488703,
        "sampled_tokens_per_second": 1090.6351582605357
      }
    },
    "designed_ceiling": {
      "counts": {
        "completions": 144,
        "injected_tokens": 288,
        "logical_model_input_tokens": 283800,
        "requests": 12,
        "sampled_tokens": 37383,
        "stage1_logical_prompt_tokens": 127932,
        "stage2_logical_prompt_tokens": 155868,
        "unique_input_prompt_tokens": 10661
      },
      "engine": {
        "adapter": null,
        "enable_prefix_caching": false,
        "enforce_eager": false,
        "gpu_memory_utilization": 0.9,
        "max_model_len": 16384,
        "max_num_batched_tokens": 32768,
        "max_num_seqs": 128
      },
      "model": "Qwen/Qwen3.5-4B",
      "model_revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a",
      "runner_sha256": "022f956ca5185fb67d5913b21465d1e465f6a1fee09067846731c87718d4527c",
      "sampling": {
        "allow_custom_prompts": false,
        "answer_max_tokens": 128,
        "frequency_penalty": 0.0,
        "greedy": false,
        "logprob_token_ids": [],
        "logprobs": null,
        "max_tokens": 128,
        "min_p": 0.0,
        "n": 12,
        "presence_penalty": 0.0,
        "prompt_logprobs": null,
        "repetition_penalty": 1.0,
        "run_seed": 2701,
        "shuffle_thinking": false,
        "temperature": 0.6,
        "thinking": "budget",
        "thinking_budget": 192,
        "top_k": 20,
        "top_p": 0.95
      },
      "schema_version": 2,
      "timing": {
        "generation_seconds": 38.00302366912365,
        "model_load_seconds": 17.697869495488703,
        "sampled_tokens_per_second": 983.6848858521907
      }
    },
    "mined": {
      "counts": {
        "completions": 144,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284088,
        "requests": 12,
        "sampled_tokens": 37369,
        "stage1_logical_prompt_tokens": 128076,
        "stage2_logical_prompt_tokens": 156012,
        "unique_input_prompt_tokens": 10673
      },
      "engine": {
        "adapter": null,
        "enable_prefix_caching": false,
        "enforce_eager": false,
        "gpu_memory_utilization": 0.9,
        "max_model_len": 16384,
        "max_num_batched_tokens": 32768,
        "max_num_seqs": 128
      },
      "model": "Qwen/Qwen3.5-4B",
      "model_revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a",
      "runner_sha256": "022f956ca5185fb67d5913b21465d1e465f6a1fee09067846731c87718d4527c",
      "sampling": {
        "allow_custom_prompts": false,
        "answer_max_tokens": 128,
        "frequency_penalty": 0.0,
        "greedy": false,
        "logprob_token_ids": [],
        "logprobs": null,
        "max_tokens": 128,
        "min_p": 0.0,
        "n": 12,
        "presence_penalty": 0.0,
        "prompt_logprobs": null,
        "repetition_penalty": 1.0,
        "run_seed": 2701,
        "shuffle_thinking": false,
        "temperature": 0.6,
        "thinking": "budget",
        "thinking_budget": 192,
        "top_k": 20,
        "top_p": 0.95
      },
      "schema_version": 2,
      "timing": {
        "generation_seconds": 37.87069963198155,
        "model_load_seconds": 17.697869495488703,
        "sampled_tokens_per_second": 986.7523009382729
      }
    },
    "mined_hint": {
      "counts": {
        "completions": 144,
        "injected_tokens": 288,
        "logical_model_input_tokens": 285528,
        "requests": 12,
        "sampled_tokens": 38039,
        "stage1_logical_prompt_tokens": 128796,
        "stage2_logical_prompt_tokens": 156732,
        "unique_input_prompt_tokens": 10733
      },
      "engine": {
        "adapter": null,
        "enable_prefix_caching": false,
        "enforce_eager": false,
        "gpu_memory_utilization": 0.9,
        "max_model_len": 16384,
        "max_num_batched_tokens": 32768,
        "max_num_seqs": 128
      },
      "model": "Qwen/Qwen3.5-4B",
      "model_revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a",
      "runner_sha256": "022f956ca5185fb67d5913b21465d1e465f6a1fee09067846731c87718d4527c",
      "sampling": {
        "allow_custom_prompts": false,
        "answer_max_tokens": 128,
        "frequency_penalty": 0.0,
        "greedy": false,
        "logprob_token_ids": [],
        "logprobs": null,
        "max_tokens": 128,
        "min_p": 0.0,
        "n": 12,
        "presence_penalty": 0.0,
        "prompt_logprobs": null,
        "repetition_penalty": 1.0,
        "run_seed": 2701,
        "shuffle_thinking": false,
        "temperature": 0.6,
        "thinking": "budget",
        "thinking_budget": 192,
        "top_k": 20,
        "top_p": 0.95
      },
      "schema_version": 2,
      "timing": {
        "generation_seconds": 38.379088904708624,
        "model_load_seconds": 17.697869495488703,
        "sampled_tokens_per_second": 991.1386925949954
      }
    },
    "random_0": {
      "counts": {
        "completions": 144,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284376,
        "requests": 12,
        "sampled_tokens": 37823,
        "stage1_logical_prompt_tokens": 128220,
        "stage2_logical_prompt_tokens": 156156,
        "unique_input_prompt_tokens": 10685
      },
      "engine": {
        "adapter": null,
        "enable_prefix_caching": false,
        "enforce_eager": false,
        "gpu_memory_utilization": 0.9,
        "max_model_len": 16384,
        "max_num_batched_tokens": 32768,
        "max_num_seqs": 128
      },
      "model": "Qwen/Qwen3.5-4B",
      "model_revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a",
      "runner_sha256": "022f956ca5185fb67d5913b21465d1e465f6a1fee09067846731c87718d4527c",
      "sampling": {
        "allow_custom_prompts": false,
        "answer_max_tokens": 128,
        "frequency_penalty": 0.0,
        "greedy": false,
        "logprob_token_ids": [],
        "logprobs": null,
        "max_tokens": 128,
        "min_p": 0.0,
        "n": 12,
        "presence_penalty": 0.0,
        "prompt_logprobs": null,
        "repetition_penalty": 1.0,
        "run_seed": 2701,
        "shuffle_thinking": false,
        "temperature": 0.6,
        "thinking": "budget",
        "thinking_budget": 192,
        "top_k": 20,
        "top_p": 0.95
      },
      "schema_version": 2,
      "timing": {
        "generation_seconds": 38.35911696776748,
        "model_load_seconds": 17.697869495488703,
        "sampled_tokens_per_second": 986.0237406346458
      }
    },
    "random_1": {
      "counts": {
        "completions": 144,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284088,
        "requests": 12,
        "sampled_tokens": 36931,
        "stage1_logical_prompt_tokens": 128076,
        "stage2_logical_prompt_tokens": 156012,
        "unique_input_prompt_tokens": 10673
      },
      "engine": {
        "adapter": null,
        "enable_prefix_caching": false,
        "enforce_eager": false,
        "gpu_memory_utilization": 0.9,
        "max_model_len": 16384,
        "max_num_batched_tokens": 32768,
        "max_num_seqs": 128
      },
      "model": "Qwen/Qwen3.5-4B",
      "model_revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a",
      "runner_sha256": "022f956ca5185fb67d5913b21465d1e465f6a1fee09067846731c87718d4527c",
      "sampling": {
        "allow_custom_prompts": false,
        "answer_max_tokens": 128,
        "frequency_penalty": 0.0,
        "greedy": false,
        "logprob_token_ids": [],
        "logprobs": null,
        "max_tokens": 128,
        "min_p": 0.0,
        "n": 12,
        "presence_penalty": 0.0,
        "prompt_logprobs": null,
        "repetition_penalty": 1.0,
        "run_seed": 2701,
        "shuffle_thinking": false,
        "temperature": 0.6,
        "thinking": "budget",
        "thinking_budget": 192,
        "top_k": 20,
        "top_p": 0.95
      },
      "schema_version": 2,
      "timing": {
        "generation_seconds": 38.153216446749866,
        "model_load_seconds": 17.697869495488703,
        "sampled_tokens_per_second": 967.9655724844141
      }
    },
    "random_2": {
      "counts": {
        "completions": 144,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284088,
        "requests": 12,
        "sampled_tokens": 37082,
        "stage1_logical_prompt_tokens": 128076,
        "stage2_logical_prompt_tokens": 156012,
        "unique_input_prompt_tokens": 10673
      },
      "engine": {
        "adapter": null,
        "enable_prefix_caching": false,
        "enforce_eager": false,
        "gpu_memory_utilization": 0.9,
        "max_model_len": 16384,
        "max_num_batched_tokens": 32768,
        "max_num_seqs": 128
      },
      "model": "Qwen/Qwen3.5-4B",
      "model_revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a",
      "runner_sha256": "022f956ca5185fb67d5913b21465d1e465f6a1fee09067846731c87718d4527c",
      "sampling": {
        "allow_custom_prompts": false,
        "answer_max_tokens": 128,
        "frequency_penalty": 0.0,
        "greedy": false,
        "logprob_token_ids": [],
        "logprobs": null,
        "max_tokens": 128,
        "min_p": 0.0,
        "n": 12,
        "presence_penalty": 0.0,
        "prompt_logprobs": null,
        "repetition_penalty": 1.0,
        "run_seed": 2701,
        "shuffle_thinking": false,
        "temperature": 0.6,
        "thinking": "budget",
        "thinking_budget": 192,
        "top_k": 20,
        "top_p": 0.95
      },
      "schema_version": 2,
      "timing": {
        "generation_seconds": 38.048863338306546,
        "model_load_seconds": 17.697869495488703,
        "sampled_tokens_per_second": 974.5889034920752
      }
    },
    "random_3": {
      "counts": {
        "completions": 144,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284088,
        "requests": 12,
        "sampled_tokens": 37675,
        "stage1_logical_prompt_tokens": 128076,
        "stage2_logical_prompt_tokens": 156012,
        "unique_input_prompt_tokens": 10673
      },
      "engine": {
        "adapter": null,
        "enable_prefix_caching": false,
        "enforce_eager": false,
        "gpu_memory_utilization": 0.9,
        "max_model_len": 16384,
        "max_num_batched_tokens": 32768,
        "max_num_seqs": 128
      },
      "model": "Qwen/Qwen3.5-4B",
      "model_revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a",
      "runner_sha256": "022f956ca5185fb67d5913b21465d1e465f6a1fee09067846731c87718d4527c",
      "sampling": {
        "allow_custom_prompts": false,
        "answer_max_tokens": 128,
        "frequency_penalty": 0.0,
        "greedy": false,
        "logprob_token_ids": [],
        "logprobs": null,
        "max_tokens": 128,
        "min_p": 0.0,
        "n": 12,
        "presence_penalty": 0.0,
        "prompt_logprobs": null,
        "repetition_penalty": 1.0,
        "run_seed": 2701,
        "shuffle_thinking": false,
        "temperature": 0.6,
        "thinking": "budget",
        "thinking_budget": 192,
        "top_k": 20,
        "top_p": 0.95
      },
      "schema_version": 2,
      "timing": {
        "generation_seconds": 37.90050444472581,
        "model_load_seconds": 17.697869495488703,
        "sampled_tokens_per_second": 994.0500938436139
      }
    },
    "random_4": {
      "counts": {
        "completions": 144,
        "injected_tokens": 288,
        "logical_model_input_tokens": 284376,
        "requests": 12,
        "sampled_tokens": 37323,
        "stage1_logical_prompt_tokens": 128220,
        "stage2_logical_prompt_tokens": 156156,
        "unique_input_prompt_tokens": 10685
      },
      "engine": {
        "adapter": null,
        "enable_prefix_caching": false,
        "enforce_eager": false,
        "gpu_memory_utilization": 0.9,
        "max_model_len": 16384,
        "max_num_batched_tokens": 32768,
        "max_num_seqs": 128
      },
      "model": "Qwen/Qwen3.5-4B",
      "model_revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a",
      "runner_sha256": "022f956ca5185fb67d5913b21465d1e465f6a1fee09067846731c87718d4527c",
      "sampling": {
        "allow_custom_prompts": false,
        "answer_max_tokens": 128,
        "frequency_penalty": 0.0,
        "greedy": false,
        "logprob_token_ids": [],
        "logprobs": null,
        "max_tokens": 128,
        "min_p": 0.0,
        "n": 12,
        "presence_penalty": 0.0,
        "prompt_logprobs": null,
        "repetition_penalty": 1.0,
        "run_seed": 2701,
        "shuffle_thinking": false,
        "temperature": 0.6,
        "thinking": "budget",
        "thinking_budget": 192,
        "top_k": 20,
        "top_p": 0.95
      },
      "schema_version": 2,
      "timing": {
        "generation_seconds": 37.87951398547739,
        "model_load_seconds": 17.697869495488703,
        "sampled_tokens_per_second": 985.3083124115386
      }
    }
  },
  "schema_version": 1,
  "selection_rule": "earliest valid sample among maximal visible-example score; hidden labels grade only",
  "smoke_gate": {
    "gates": {
      "answer_truncation": false,
      "base_parse": true,
      "designed_oracle_not_below_base": false,
      "designed_parse": true,
      "designed_valid_macro_candidates": false,
      "overall_parse": true
    },
    "metrics": {
      "base_oracle_hidden_coverage": 0.08333333333333333,
      "base_parse_rate": 0.6111111111111112,
      "designed_oracle_hidden_coverage": 0.0,
      "designed_parse_rate": 0.5694444444444444,
      "designed_valid_macro_using_candidates": 0,
      "overall_parse_rate": 0.5972222222222222,
      "overall_truncation_rate": 0.40046296296296297
    },
    "pass": false,
    "reasons": [
      "designed_valid_macro_candidates",
      "answer_truncation",
      "designed_oracle_not_below_base"
    ],
    "required_arms_present": true
  }
}
