{
  "run": "pilot_expand_copy_seed123",
  "arm": "expand_copy",
  "seed": 123,
  "args": {
    "run_name": "pilot_expand_copy_seed123",
    "arm": "expand_copy",
    "model_id": "Qwen/Qwen3-4B",
    "seed": 123,
    "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "stage": "stage1_max8",
      "stage_idx": 1,
      "local_step": 1,
      "global_step": 1,
      "compiler_max_steps": 8,
      "elapsed_sec": 1.055,
      "expansion_event": "init",
      "init_loss": 4.492895126342773,
      "op_loss": 1.137392520904541,
      "arg_loss": 4.5463128089904785,
      "state_loss": 4.580261707305908,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.581772804260254,
      "loss": 46.45625686645508
    },
    {
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "stage": "stage1_max8",
      "stage_idx": 1,
      "local_step": 2,
      "global_step": 2,
      "compiler_max_steps": 8,
      "elapsed_sec": 1.49,
      "expansion_event": "init",
      "init_loss": 4.760336875915527,
      "op_loss": 1.1789627075195312,
      "arg_loss": 4.521578788757324,
      "state_loss": 4.5789265632629395,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.579809188842773,
      "loss": 47.465362548828125
    },
    {
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "stage": "stage1_max8",
      "stage_idx": 1,
      "local_step": 4,
      "global_step": 4,
      "compiler_max_steps": 8,
      "elapsed_sec": 2.359,
      "expansion_event": "init",
      "init_loss": 4.8263959884643555,
      "op_loss": 1.2378901243209839,
      "arg_loss": 4.554002285003662,
      "state_loss": 4.579301834106445,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.580705642700195,
      "loss": 47.91948699951172
    },
    {
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "stage": "stage2_max16",
      "stage_idx": 2,
      "local_step": 1,
      "global_step": 5,
      "compiler_max_steps": 16,
      "elapsed_sec": 4.921,
      "expansion_event": "expand_8_to_16_copy_last",
      "init_loss": 4.641975402832031,
      "op_loss": 1.0597169399261475,
      "arg_loss": 4.434762954711914,
      "state_loss": 4.5796356201171875,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.579626083374023,
      "loss": 46.52593231201172
    },
    {
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "stage": "stage2_max16",
      "stage_idx": 2,
      "local_step": 2,
      "global_step": 6,
      "compiler_max_steps": 16,
      "elapsed_sec": 5.408,
      "expansion_event": "expand_8_to_16_copy_last",
      "init_loss": 4.393392562866211,
      "op_loss": 1.2387518882751465,
      "arg_loss": 4.643195152282715,
      "state_loss": 4.578445911407471,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.576048851013184,
      "loss": 46.53959655761719
    },
    {
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "stage": "stage3_max24",
      "stage_idx": 3,
      "local_step": 1,
      "global_step": 7,
      "compiler_max_steps": 24,
      "elapsed_sec": 13.177,
      "expansion_event": "expand_16_to_24_copy_last",
      "init_loss": 4.705228805541992,
      "op_loss": 1.0877423286437988,
      "arg_loss": 4.307279586791992,
      "state_loss": 4.580263614654541,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.580606460571289,
      "loss": 46.298648834228516
    },
    {
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "stage": "stage3_max24",
      "stage_idx": 3,
      "local_step": 2,
      "global_step": 8,
      "compiler_max_steps": 24,
      "elapsed_sec": 13.99,
      "expansion_event": "expand_16_to_24_copy_last",
      "init_loss": 4.449899673461914,
      "op_loss": 1.0725171566009521,
      "arg_loss": 4.471164703369141,
      "state_loss": 4.581061363220215,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.581439971923828,
      "loss": 45.91927719116211
    },
    {
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "stage": "stage3_max24",
      "stage_idx": 3,
      "local_step": 4,
      "global_step": 10,
      "compiler_max_steps": 24,
      "elapsed_sec": 15.512,
      "expansion_event": "expand_16_to_24_copy_last",
      "init_loss": 4.100093841552734,
      "op_loss": 1.0793756246566772,
      "arg_loss": 4.4627299308776855,
      "state_loss": 4.579633712768555,
      "state_train_accuracy": 0.0,
      "executor_loss": 4.578847885131836,
      "loss": 44.489158630371094
    }
  ],
  "metrics": [
    {
      "stage": "stage1_max8",
      "split": "standard_L8",
      "n": 8,
      "max_steps": 8,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010261820629239082,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.375,
      "arg_accuracy": 0.109375,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 4,
      "stage_idx": 1
    },
    {
      "stage": "stage1_max8",
      "split": "paraphrase_L8",
      "n": 8,
      "max_steps": 8,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010263125877827406,
      "direct_accuracy": NaN,
      "init_accuracy": 0.125,
      "op_accuracy": 0.34375,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 4,
      "stage_idx": 1
    },
    {
      "stage": "stage1_max8",
      "split": "heldout_L8",
      "n": 8,
      "max_steps": 8,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.01026088371872902,
      "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,
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 4,
      "stage_idx": 1
    },
    {
      "stage": "stage1_max8",
      "split": "paired_L8",
      "n": 8,
      "max_steps": 8,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010258609429001808,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.359375,
      "arg_accuracy": 0.03125,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "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": 0.010260766372084618,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.40625,
      "arg_accuracy": 0.0625,
      "program_exact": 0.0,
      "state_accuracy": 0.03125,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 4,
      "stage_idx": 1
    },
    {
      "stage": "stage2_max16",
      "split": "standard_L8",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.125,
      "executor_target_mass": 0.010893609374761581,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3125,
      "arg_accuracy": 0.015625,
      "program_exact": 0.0,
      "state_accuracy": 0.03125,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "paraphrase_L8",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010258875321596861,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.328125,
      "arg_accuracy": 0.078125,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "heldout_L8",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010253534652292728,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.296875,
      "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,
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "paired_L8",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010255812667310238,
      "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,
      "executor_pair_answer_consistency": 0.25,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 0.0,
      "compiler_pair_state_consistency": 0.0,
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "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": 0.010253304615616798,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.265625,
      "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,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "standard_L16",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010254469700157642,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3046875,
      "arg_accuracy": 0.0546875,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "paraphrase_L16",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010254162829369307,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3515625,
      "arg_accuracy": 0.0390625,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "heldout_L16",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010257975198328495,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.328125,
      "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,
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage2_max16",
      "split": "paired_L16",
      "n": 8,
      "max_steps": 16,
      "executor_accuracy": 0.125,
      "executor_target_mass": 0.010260019451379776,
      "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,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "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": 0.010253910440951586,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.390625,
      "arg_accuracy": 0.0546875,
      "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": 0.25,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 0.25,
      "compiler_pair_state_consistency": 0.25,
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 6,
      "stage_idx": 2
    },
    {
      "stage": "stage3_max24",
      "split": "standard_L8",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010264729615300894,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.265625,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paraphrase_L8",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010267951525747776,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "heldout_L8",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.125,
      "executor_target_mass": 0.010274017229676247,
      "direct_accuracy": NaN,
      "init_accuracy": 0.125,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paired_L8",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010267829522490501,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.4375,
      "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": 1.0,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 1.0,
      "compiler_pair_state_consistency": 1.0,
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "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": 0.010271389968693256,
      "direct_accuracy": NaN,
      "init_accuracy": 0.25,
      "op_accuracy": 0.34375,
      "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,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "standard_L16",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010273633990436792,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3828125,
      "arg_accuracy": 0.078125,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paraphrase_L16",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.0102724670432508,
      "direct_accuracy": NaN,
      "init_accuracy": 0.125,
      "op_accuracy": 0.3203125,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "heldout_L16",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.01026745093986392,
      "direct_accuracy": NaN,
      "init_accuracy": 0.125,
      "op_accuracy": 0.3671875,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paired_L16",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.01026938110589981,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.21875,
      "arg_accuracy": 0.0703125,
      "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": 0.75,
      "compiler_pair_state_consistency": 0.75,
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "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": 0.010279977694153786,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.265625,
      "arg_accuracy": 0.078125,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "standard_L24",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010267865844070911,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3802083333333333,
      "arg_accuracy": 0.06770833333333333,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paraphrase_L24",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010768687818199396,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.34375,
      "arg_accuracy": 0.057291666666666664,
      "program_exact": 0.0,
      "state_accuracy": 0.005208333333333333,
      "state_all_exact": 0.0,
      "state_prefix_fraction": 0.0,
      "state_mean_correct_prefix": 0.0,
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "heldout_L24",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.01026909751817584,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.296875,
      "arg_accuracy": 0.046875,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    },
    {
      "stage": "stage3_max24",
      "split": "paired_L24",
      "n": 8,
      "max_steps": 24,
      "executor_accuracy": 0.0,
      "executor_target_mass": 0.010267456993460655,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.2916666666666667,
      "arg_accuracy": 0.07291666666666667,
      "program_exact": 0.0,
      "state_accuracy": 0.020833333333333332,
      "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_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "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": 0.010271199513226748,
      "direct_accuracy": NaN,
      "init_accuracy": 0.0,
      "op_accuracy": 0.3125,
      "arg_accuracy": 0.020833333333333332,
      "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": 0.5,
      "executor_pair_both_correct": 0.0,
      "compiler_pair_program_consistency": 0.5,
      "compiler_pair_state_consistency": 0.5,
      "run": "pilot_expand_copy_seed123",
      "arm": "expand_copy",
      "seed": 123,
      "global_step": 10,
      "stage_idx": 3
    }
  ],
  "checkpoints": [],
  "elapsed_sec": 30.605
}