{
  "run": "pilot_expand_copy_seed456",
  "arm": "expand_copy",
  "seed": 456,
  "args": {
    "run_name": "pilot_expand_copy_seed456",
    "arm": "expand_copy",
    "model_id": "Qwen/Qwen3-4B",
    "seed": 456,
    "modulus": 97,
    "stage_max_steps": "8,16,24",
    "stage_min_lengths": "1,1,8",
    "stage_train_max_lengths": "8,16,24",
    "stage_steps": "4,2,4",
    "train_examples": 32,
    "train_template_mode": "mixed",
    "eval_lengths": "8,16,24",
    "eval_examples": 8,
    "paired_eval_pairs": 4,
    "batch_size": 2,
    "grad_accum": 1,
    "eval_batch_size": 4,
    "max_length": 2048,
    "head_width": 192,
    "compiler_layers": 1,
    "compiler_heads": 4,
    "compiler_dropout": 0.05,
    "expansion_noise": 0.005,
    "expansion_mode": "copy_last",
    "trace_loss_weight": 1.0,
    "executor_loss_weight": 1.0,
    "state_loss_weight": 1.0,
    "init_trace_loss_weight": 4.0,
    "op_trace_loss_weight": 1.0,
    "arg_trace_loss_weight": 4.0,
    "direct_head_weight": 0.0,
    "lr": 0.0002,
    "weight_decay": 0.0,
    "optimizer": "paged_adamw_8bit",
    "max_grad_norm": 1.0,
    "torch_dtype": "bf16",
    "load_in_4bit": true,
    "device_map": "auto",
    "use_lora": true,
    "gradient_checkpointing": true,
    "lora_r": 4,
    "lora_alpha": 8,
    "lora_dropout": 0.05,
    "lora_target_modules": "all-linear",
    "log_interval": 2,
    "save_checkpoints": false,
    "checkpoint_dir": "",
    "selection_split": "paired_L24",
    "selection_metric": "executor_pair_both_correct"
  },
  "metadata": {
    "python": "3.12.3",
    "platform": "Linux-6.8.0-62-generic-x86_64-with-glibc2.39",
    "torch": "2.8.0+cu128",
    "cuda_available": true,
    "transformers_loader": "AutoModelForCausalLM",
    "peft_installed": true,
    "use_lora": true,
    "lora_r": 4,
    "model_id": "Qwen/Qwen3-4B",
    "gpu_name": "NVIDIA RTX 6000 Ada Generation",
    "gpu_vram_gb": 47.383
  },
  "stages": [
    {
      "name": "stage1_max8",
      "max_steps": 8,
      "train_min_len": 1,
      "train_max_len": 8,
      "steps": 4
    },
    {
      "name": "stage2_max16",
      "max_steps": 16,
      "train_min_len": 1,
      "train_max_len": 16,
      "steps": 2
    },
    {
      "name": "stage3_max24",
      "max_steps": 24,
      "train_min_len": 8,
      "train_max_len": 24,
      "steps": 4
    }
  ],
  "train_log": [
    {
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "stage": "stage1_max8",
      "stage_idx": 1,
      "local_step": 1,
      "global_step": 1,
      "compiler_max_steps": 8,
      "elapsed_sec": 1.094,
      "expansion_event": "init",
      "init_loss": 4.751434803009033,
      "op_loss": 1.0811601877212524,
      "arg_loss": 4.649412155151367,
      "state_loss": 4.579232692718506,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.578859329223633,
      "loss": 47.84263610839844
    },
    {
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "stage": "stage1_max8",
      "stage_idx": 1,
      "local_step": 2,
      "global_step": 2,
      "compiler_max_steps": 8,
      "elapsed_sec": 1.538,
      "expansion_event": "init",
      "init_loss": 4.725821495056152,
      "op_loss": 1.1368590593338013,
      "arg_loss": 4.527618408203125,
      "state_loss": 4.578524112701416,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.5780510902404785,
      "loss": 47.30719757080078
    },
    {
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "stage": "stage1_max8",
      "stage_idx": 1,
      "local_step": 4,
      "global_step": 4,
      "compiler_max_steps": 8,
      "elapsed_sec": 2.429,
      "expansion_event": "init",
      "init_loss": 4.776422023773193,
      "op_loss": 1.1201473474502563,
      "arg_loss": 4.608232498168945,
      "state_loss": 4.579368591308594,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.579306125640869,
      "loss": 47.81744384765625
    },
    {
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "stage": "stage2_max16",
      "stage_idx": 2,
      "local_step": 1,
      "global_step": 5,
      "compiler_max_steps": 16,
      "elapsed_sec": 5.023,
      "expansion_event": "expand_8_to_16_copy_last",
      "init_loss": 4.682049751281738,
      "op_loss": 1.0961625576019287,
      "arg_loss": 4.501832485198975,
      "state_loss": 4.579159736633301,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.579472541809082,
      "loss": 46.99032211303711
    },
    {
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "stage": "stage2_max16",
      "stage_idx": 2,
      "local_step": 2,
      "global_step": 6,
      "compiler_max_steps": 16,
      "elapsed_sec": 5.479,
      "expansion_event": "expand_8_to_16_copy_last",
      "init_loss": 4.486515522003174,
      "op_loss": 1.0839159488677979,
      "arg_loss": 4.531767845153809,
      "state_loss": 4.578577995300293,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.5789995193481445,
      "loss": 46.31462860107422
    },
    {
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "stage": "stage3_max24",
      "stage_idx": 3,
      "local_step": 1,
      "global_step": 7,
      "compiler_max_steps": 24,
      "elapsed_sec": 13.312,
      "expansion_event": "expand_16_to_24_copy_last",
      "init_loss": 4.530548095703125,
      "op_loss": 1.072959542274475,
      "arg_loss": 4.445160388946533,
      "state_loss": 4.580003261566162,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.580333232879639,
      "loss": 46.13612747192383
    },
    {
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "stage": "stage3_max24",
      "stage_idx": 3,
      "local_step": 2,
      "global_step": 8,
      "compiler_max_steps": 24,
      "elapsed_sec": 14.145,
      "expansion_event": "expand_16_to_24_copy_last",
      "init_loss": 4.636808395385742,
      "op_loss": 1.0597426891326904,
      "arg_loss": 4.404459476470947,
      "state_loss": 4.580131530761719,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.5814409255981445,
      "loss": 46.38638687133789
    },
    {
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "stage": "stage3_max24",
      "stage_idx": 3,
      "local_step": 4,
      "global_step": 10,
      "compiler_max_steps": 24,
      "elapsed_sec": 15.743,
      "expansion_event": "expand_16_to_24_copy_last",
      "init_loss": 4.304520606994629,
      "op_loss": 1.108992099761963,
      "arg_loss": 4.570666790008545,
      "state_loss": 4.57864236831665,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.579602241516113,
      "loss": 45.76798629760742
    }
  ],
  "metrics": [
    {
      "stage": "stage1_max8",
      "split": "standard_L8",
      "n": 8,
      "max_steps": 8,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010255235712975264,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.4375,
      "arg_accuracy": 0.078125,
      "program_exact": 0.0,
      "state_accuracy": 0.015625,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.015625,
      "state_mean_correct_prefix": 0.125,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 4,
      "stage_idx": 1
    },
    {
      "stage": "stage1_max8",
      "split": "paraphrase_L8",
      "n": 8,
      "max_steps": 8,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010250900406390429,
      "direct_accuracy": NaN,
      "init_accuracy": 0.125,
      "op_accuracy": 0.3125,
      "arg_accuracy": 0.0625,
      "program_exact": 0.0,
      "state_accuracy": 0.015625,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 4,
      "stage_idx": 1
    },
    {
      "stage": "stage1_max8",
      "split": "heldout_L8",
      "n": 8,
      "max_steps": 8,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010259821079671383,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.359375,
      "arg_accuracy": 0.03125,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 4,
      "stage_idx": 1
    },
    {
      "stage": "stage1_max8",
      "split": "paired_L8",
      "n": 8,
      "max_steps": 8,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.01025626016780734,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.375,
      "arg_accuracy": 0.0625,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 1.0,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 1.0,
      "compiler_pair_state_consistency": 1.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 4,
      "stage_idx": 1
    },
    {
      "stage": "stage1_max8",
      "split": "paired_heldout_L8",
      "n": 8,
      "max_steps": 8,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010250478517264128,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.28125,
      "arg_accuracy": 0.09375,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 0.5,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 0.5,
      "compiler_pair_state_consistency": 0.5,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 4,
      "stage_idx": 1
    },
    {
      "stage": "stage2_max16",
      "split": "standard_L8",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010251764673739672,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.203125,
      "arg_accuracy": 0.0625,
      "program_exact": 0.0,
      "state_accuracy": 0.015625,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "paraphrase_L8",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010972163639962673,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.265625,
      "arg_accuracy": 0.046875,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "heldout_L8",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.01025184104219079,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.40625,
      "arg_accuracy": 0.046875,
      "program_exact": 0.0,
      "state_accuracy": 0.015625,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "paired_L8",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010253188200294971,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.1875,
      "arg_accuracy": 0.046875,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 0.5,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 0.25,
      "compiler_pair_state_consistency": 0.25,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "paired_heldout_L8",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010251559782773256,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.21875,
      "arg_accuracy": 0.015625,
      "program_exact": 0.0,
      "state_accuracy": 0.015625,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 0.0,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 0.0,
      "compiler_pair_state_consistency": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "standard_L16",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.01024982612580061,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.359375,
      "arg_accuracy": 0.0703125,
      "program_exact": 0.0,
      "state_accuracy": 0.0078125,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "paraphrase_L16",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010250021703541279,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.4140625,
      "arg_accuracy": 0.0234375,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "heldout_L16",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010245527606457472,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.390625,
      "arg_accuracy": 0.046875,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "paired_L16",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010249977465718985,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.375,
      "arg_accuracy": 0.1015625,
      "program_exact": 0.0,
      "state_accuracy": 0.015625,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 0.75,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 0.25,
      "compiler_pair_state_consistency": 0.25,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "paired_heldout_L16",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010242489166557789,
      "direct_accuracy": NaN,
      "init_accuracy": 0.25,
      "op_accuracy": 0.375,
      "arg_accuracy": 0.0390625,
      "program_exact": 0.0,
      "state_accuracy": 0.0078125,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 0.0,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 0.0,
      "compiler_pair_state_consistency": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage3_max24",
      "split": "standard_L8",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010264409240335226,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.375,
      "arg_accuracy": 0.0,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paraphrase_L8",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010264917276799679,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.296875,
      "arg_accuracy": 0.046875,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "heldout_L8",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010262256022542715,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.34375,
      "arg_accuracy": 0.0625,
      "program_exact": 0.0,
      "state_accuracy": 0.015625,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paired_L8",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010268999729305506,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3125,
      "arg_accuracy": 0.0,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 1.0,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 1.0,
      "compiler_pair_state_consistency": 1.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paired_heldout_L8",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010265371762216091,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.40625,
      "arg_accuracy": 0.0625,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 1.0,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 1.0,
      "compiler_pair_state_consistency": 1.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "standard_L16",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010266337543725967,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.359375,
      "arg_accuracy": 0.0546875,
      "program_exact": 0.0,
      "state_accuracy": 0.0078125,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paraphrase_L16",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010264421347528696,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.2734375,
      "arg_accuracy": 0.0234375,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "heldout_L16",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010263933800160885,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3203125,
      "arg_accuracy": 0.0625,
      "program_exact": 0.0,
      "state_accuracy": 0.015625,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paired_L16",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.01026511937379837,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.34375,
      "arg_accuracy": 0.046875,
      "program_exact": 0.0,
      "state_accuracy": 0.015625,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 1.0,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 1.0,
      "compiler_pair_state_consistency": 1.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paired_heldout_L16",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.01026006555184722,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.25,
      "arg_accuracy": 0.09375,
      "program_exact": 0.0,
      "state_accuracy": 0.015625,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 1.0,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 1.0,
      "compiler_pair_state_consistency": 1.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "standard_L24",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.01026576105505228,
      "direct_accuracy": NaN,
      "init_accuracy": 0.25,
      "op_accuracy": 0.2916666666666667,
      "arg_accuracy": 0.06770833333333333,
      "program_exact": 0.0,
      "state_accuracy": 0.010416666666666666,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.005208333333333333,
      "state_mean_correct_prefix": 0.125,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paraphrase_L24",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010263972450047731,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3854166666666667,
      "arg_accuracy": 0.057291666666666664,
      "program_exact": 0.0,
      "state_accuracy": 0.010416666666666666,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.005208333333333333,
      "state_mean_correct_prefix": 0.125,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "heldout_L24",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010266322642564774,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3385416666666667,
      "arg_accuracy": 0.03125,
      "program_exact": 0.0,
      "state_accuracy": 0.010416666666666666,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paired_L24",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010264867916703224,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.40625,
      "arg_accuracy": 0.041666666666666664,
      "program_exact": 0.0,
      "state_accuracy": 0.010416666666666666,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 1.0,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 1.0,
      "compiler_pair_state_consistency": 1.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paired_heldout_L24",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010260398034006357,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3333333333333333,
      "arg_accuracy": 0.0625,
      "program_exact": 0.0,
      "state_accuracy": 0.0,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "executor_pair_answer_consistency": 1.0,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 1.0,
      "compiler_pair_state_consistency": 1.0,
      "run": "pilot_expand_copy_seed456",
      "arm": "expand_copy",
      "seed": 456,
      "global_step": 10,
      "stage_idx": 3
    }
  ],
  "checkpoints": [],
  "elapsed_sec": 30.932
}