{
  "mode": "joint",
  "args": {
    "mode": "joint",
    "task": "belief_line",
    "seed": 23,
    "modulus": 31,
    "op_family": "full",
    "train_min_len": 1,
    "train_max_len": 8,
    "eval_lengths": "4,8,12,16,24",
    "eval_k": "0,1,2,4,8,12,16,24",
    "transition_temperature": 1.0,
    "dim": 128,
    "heads": 4,
    "compiler_layers": 2,
    "dropout": 0.05,
    "train_steps": 500,
    "batch_size": 512,
    "eval_batch_size": 512,
    "eval_examples": 1024,
    "lr": 0.03,
    "weight_decay": 0.0,
    "max_grad_norm": 1.0,
    "log_every": 25,
    "eval_every": 250,
    "output_dir": "experiments/joint_register_executor/runs/main_belief_joint_mod31",
    "checkpoint_dir": "large_artifacts/joint_register_executor/checkpoints/main_belief_joint_mod31"
  },
  "env": {
    "python": "3.12.3",
    "torch": "2.8.0+cu128",
    "device": "cuda",
    "cuda_name": "NVIDIA RTX 6000 Ada Generation"
  },
  "train": [
    {
      "step": 1,
      "loss": 6.486381530761719,
      "pair_acc": 0.0390625,
      "grad_norm": 0.004529037978500128,
      "transition_entropy": 3.4339387118816376
    },
    {
      "step": 25,
      "loss": 6.2788896560668945,
      "pair_acc": 0.2109375,
      "grad_norm": 0.004409412853419781,
      "transition_entropy": 3.4159174859523773
    },
    {
      "step": 50,
      "loss": 5.943323135375977,
      "pair_acc": 0.43359375,
      "grad_norm": 0.007084026467055082,
      "transition_entropy": 3.3289737701416016
    },
    {
      "step": 75,
      "loss": 5.196156978607178,
      "pair_acc": 0.8046875,
      "grad_norm": 0.011927195824682713,
      "transition_entropy": 2.9113211631774902
    },
    {
      "step": 100,
      "loss": 4.017120838165283,
      "pair_acc": 1.0,
      "grad_norm": 0.0073243798688054085,
      "transition_entropy": 2.210595652461052
    },
    {
      "step": 125,
      "loss": 3.6739721298217773,
      "pair_acc": 1.0,
      "grad_norm": 0.003243634244427085,
      "transition_entropy": 1.7925750091671944
    },
    {
      "step": 150,
      "loss": 3.5736560821533203,
      "pair_acc": 1.0,
      "grad_norm": 0.001969551434740424,
      "transition_entropy": 1.7109892293810844
    },
    {
      "step": 175,
      "loss": 3.5385866165161133,
      "pair_acc": 1.0,
      "grad_norm": 0.0014448541915044188,
      "transition_entropy": 1.596084713935852
    },
    {
      "step": 200,
      "loss": 3.5106143951416016,
      "pair_acc": 1.0,
      "grad_norm": 0.001072369865141809,
      "transition_entropy": 1.6395873576402664
    },
    {
      "step": 225,
      "loss": 3.4929776191711426,
      "pair_acc": 1.0,
      "grad_norm": 0.00083963229553774,
      "transition_entropy": 1.6488552298396826
    },
    {
      "step": 250,
      "loss": 3.48574161529541,
      "pair_acc": 1.0,
      "grad_norm": 0.000729018822312355,
      "transition_entropy": 1.5831627659499645
    },
    {
      "step": 275,
      "loss": 3.478386640548706,
      "pair_acc": 1.0,
      "grad_norm": 0.0006282042013481259,
      "transition_entropy": 1.5455889608711004
    },
    {
      "step": 300,
      "loss": 3.4716906547546387,
      "pair_acc": 1.0,
      "grad_norm": 0.0005199888837523758,
      "transition_entropy": 1.5670911967754364
    },
    {
      "step": 325,
      "loss": 3.466301202774048,
      "pair_acc": 1.0,
      "grad_norm": 0.0004569691955111921,
      "transition_entropy": 1.5970662292093039
    },
    {
      "step": 350,
      "loss": 3.462493658065796,
      "pair_acc": 1.0,
      "grad_norm": 0.0004071654402650893,
      "transition_entropy": 1.5738452998921275
    },
    {
      "step": 375,
      "loss": 3.46008038520813,
      "pair_acc": 1.0,
      "grad_norm": 0.00037579942727461457,
      "transition_entropy": 1.4897717367857695
    },
    {
      "step": 400,
      "loss": 3.457951784133911,
      "pair_acc": 1.0,
      "grad_norm": 0.0003484499757178128,
      "transition_entropy": 1.4750592466443777
    },
    {
      "step": 425,
      "loss": 3.455048084259033,
      "pair_acc": 1.0,
      "grad_norm": 0.00030259665800258517,
      "transition_entropy": 1.5441008377820253
    },
    {
      "step": 450,
      "loss": 3.452547788619995,
      "pair_acc": 1.0,
      "grad_norm": 0.00027222433709539473,
      "transition_entropy": 1.603704845532775
    },
    {
      "step": 475,
      "loss": 3.451488971710205,
      "pair_acc": 1.0,
      "grad_norm": 0.00024706972180865705,
      "transition_entropy": 1.554142584092915
    },
    {
      "step": 500,
      "loss": 3.450493097305298,
      "pair_acc": 1.0,
      "grad_norm": 0.00022936986351851374,
      "transition_entropy": 1.446831385139376
    }
  ],
  "eval": [
    {
      "step": 250,
      "rows": [
        {
          "model": "joint",
          "length": 4,
          "k": 0,
          "n": 1024,
          "pair_accuracy": 0.080078125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0423702470703125,
          "target_nll": 19.990714073242188,
          "top1_on_support": 0.080078125
        },
        {
          "model": "joint",
          "length": 4,
          "k": 1,
          "n": 1024,
          "pair_accuracy": 0.08203125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0790061650390625,
          "target_nll": 10.23883056640625,
          "top1_on_support": 0.08203125
        },
        {
          "model": "joint",
          "length": 4,
          "k": 2,
          "n": 1024,
          "pair_accuracy": 0.16015625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.1499190029296875,
          "target_nll": 9.157584666992188,
          "top1_on_support": 0.16015625
        },
        {
          "model": "joint",
          "length": 4,
          "k": 4,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9316797255859375,
          "target_nll": 3.50494909375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 4,
          "k": 8,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9316797255859375,
          "target_nll": 3.50494909375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 4,
          "k": 12,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9316797255859375,
          "target_nll": 3.50494909375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 4,
          "k": 16,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9316797255859375,
          "target_nll": 3.50494909375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 4,
          "k": 24,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9316797255859375,
          "target_nll": 3.50494909375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 8,
          "k": 0,
          "n": 1024,
          "pair_accuracy": 0.033203125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.030777501953125,
          "target_nll": 20.191145896484375,
          "top1_on_support": 0.033203125
        },
        {
          "model": "joint",
          "length": 8,
          "k": 1,
          "n": 1024,
          "pair_accuracy": 0.0400390625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.03817937109375,
          "target_nll": 10.544524192382813,
          "top1_on_support": 0.0400390625
        },
        {
          "model": "joint",
          "length": 8,
          "k": 2,
          "n": 1024,
          "pair_accuracy": 0.0390625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0404049736328125,
          "target_nll": 9.935679913085938,
          "top1_on_support": 0.0390625
        },
        {
          "model": "joint",
          "length": 8,
          "k": 4,
          "n": 1024,
          "pair_accuracy": 0.0517578125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.055590390625,
          "target_nll": 9.22275209375,
          "top1_on_support": 0.0517578125
        },
        {
          "model": "joint",
          "length": 8,
          "k": 8,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.8707091806640626,
          "target_nll": 3.5729387998046875,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 8,
          "k": 12,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.8707091806640626,
          "target_nll": 3.5729387998046875,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 8,
          "k": 16,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.8707091806640626,
          "target_nll": 3.5729387998046875,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 8,
          "k": 24,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.8707091806640626,
          "target_nll": 3.5729387998046875,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 12,
          "k": 0,
          "n": 1024,
          "pair_accuracy": 0.03125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.032069083984375,
          "target_nll": 20.16881370605469,
          "top1_on_support": 0.03125
        },
        {
          "model": "joint",
          "length": 12,
          "k": 1,
          "n": 1024,
          "pair_accuracy": 0.0361328125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0343538740234375,
          "target_nll": 10.579910754882812,
          "top1_on_support": 0.0361328125
        },
        {
          "model": "joint",
          "length": 12,
          "k": 2,
          "n": 1024,
          "pair_accuracy": 0.0302734375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0317498515625,
          "target_nll": 9.990264892578125,
          "top1_on_support": 0.0302734375
        },
        {
          "model": "joint",
          "length": 12,
          "k": 4,
          "n": 1024,
          "pair_accuracy": 0.029296875,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0343153720703125,
          "target_nll": 9.371683120117188,
          "top1_on_support": 0.029296875
        },
        {
          "model": "joint",
          "length": 12,
          "k": 8,
          "n": 1024,
          "pair_accuracy": 0.0546875,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.052645560546875,
          "target_nll": 8.61769533203125,
          "top1_on_support": 0.0546875
        },
        {
          "model": "joint",
          "length": 12,
          "k": 12,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.8123480087890625,
          "target_nll": 3.6427049638671876,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 12,
          "k": 16,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.8123479482421875,
          "target_nll": 3.6427049638671876,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 12,
          "k": 24,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.8123479482421875,
          "target_nll": 3.6427049638671876,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 16,
          "k": 0,
          "n": 1024,
          "pair_accuracy": 0.0400390625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0316280556640625,
          "target_nll": 20.17644023925781,
          "top1_on_support": 0.0400390625
        },
        {
          "model": "joint",
          "length": 16,
          "k": 1,
          "n": 1024,
          "pair_accuracy": 0.0419921875,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0328432421875,
          "target_nll": 10.596554279296875,
          "top1_on_support": 0.0419921875
        },
        {
          "model": "joint",
          "length": 16,
          "k": 2,
          "n": 1024,
          "pair_accuracy": 0.02734375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0334617421875,
          "target_nll": 9.972254752929688,
          "top1_on_support": 0.02734375
        },
        {
          "model": "joint",
          "length": 16,
          "k": 4,
          "n": 1024,
          "pair_accuracy": 0.0234375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.03151291015625,
          "target_nll": 9.372495174804687,
          "top1_on_support": 0.0234375
        },
        {
          "model": "joint",
          "length": 16,
          "k": 8,
          "n": 1024,
          "pair_accuracy": 0.037109375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0347317724609375,
          "target_nll": 8.71758985546875,
          "top1_on_support": 0.037109375
        },
        {
          "model": "joint",
          "length": 16,
          "k": 12,
          "n": 1024,
          "pair_accuracy": 0.064453125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0504943876953125,
          "target_nll": 8.257741928710937,
          "top1_on_support": 0.064453125
        },
        {
          "model": "joint",
          "length": 16,
          "k": 16,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.7560354169921875,
          "target_nll": 3.7150186298828123,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 16,
          "k": 24,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.7560354169921875,
          "target_nll": 3.7150186298828123,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 24,
          "k": 0,
          "n": 1024,
          "pair_accuracy": 0.03515625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0343687314453125,
          "target_nll": 20.1290550234375,
          "top1_on_support": 0.03515625
        },
        {
          "model": "joint",
          "length": 24,
          "k": 1,
          "n": 1024,
          "pair_accuracy": 0.02734375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0311171123046875,
          "target_nll": 10.605134963867188,
          "top1_on_support": 0.02734375
        },
        {
          "model": "joint",
          "length": 24,
          "k": 2,
          "n": 1024,
          "pair_accuracy": 0.037109375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0322392646484375,
          "target_nll": 9.985058784179687,
          "top1_on_support": 0.037109375
        },
        {
          "model": "joint",
          "length": 24,
          "k": 4,
          "n": 1024,
          "pair_accuracy": 0.0400390625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.03300662109375,
          "target_nll": 9.365402221679688,
          "top1_on_support": 0.0400390625
        },
        {
          "model": "joint",
          "length": 24,
          "k": 8,
          "n": 1024,
          "pair_accuracy": 0.02734375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.031170353515625,
          "target_nll": 8.748178958984376,
          "top1_on_support": 0.02734375
        },
        {
          "model": "joint",
          "length": 24,
          "k": 12,
          "n": 1024,
          "pair_accuracy": 0.0390625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0345567490234375,
          "target_nll": 8.357385159179687,
          "top1_on_support": 0.0390625
        },
        {
          "model": "joint",
          "length": 24,
          "k": 16,
          "n": 1024,
          "pair_accuracy": 0.025390625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0353228857421875,
          "target_nll": 8.09844970703125,
          "top1_on_support": 0.025390625
        },
        {
          "model": "joint",
          "length": 24,
          "k": 24,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.6526665986328125,
          "target_nll": 3.862830638671875,
          "top1_on_support": 1.0
        }
      ]
    },
    {
      "step": 500,
      "rows": [
        {
          "model": "joint",
          "length": 4,
          "k": 0,
          "n": 1024,
          "pair_accuracy": 0.07421875,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0437563359375,
          "target_nll": 19.966751098632812,
          "top1_on_support": 0.07421875
        },
        {
          "model": "joint",
          "length": 4,
          "k": 1,
          "n": 1024,
          "pair_accuracy": 0.080078125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.079072169921875,
          "target_nll": 11.302562713867188,
          "top1_on_support": 0.080078125
        },
        {
          "model": "joint",
          "length": 4,
          "k": 2,
          "n": 1024,
          "pair_accuracy": 0.142578125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.1451847412109375,
          "target_nll": 10.166954516601562,
          "top1_on_support": 0.142578125
        },
        {
          "model": "joint",
          "length": 4,
          "k": 4,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.977619587890625,
          "target_nll": 3.4566398857421876,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 4,
          "k": 8,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9776196484375,
          "target_nll": 3.4566397666015627,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 4,
          "k": 12,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9776196484375,
          "target_nll": 3.4566397666015627,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 4,
          "k": 16,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9776196484375,
          "target_nll": 3.4566397666015627,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 4,
          "k": 24,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9776196484375,
          "target_nll": 3.4566397666015627,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 8,
          "k": 0,
          "n": 1024,
          "pair_accuracy": 0.033203125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.030998015625,
          "target_nll": 20.187332153320312,
          "top1_on_support": 0.033203125
        },
        {
          "model": "joint",
          "length": 8,
          "k": 1,
          "n": 1024,
          "pair_accuracy": 0.0361328125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0378911357421875,
          "target_nll": 11.6549882890625,
          "top1_on_support": 0.0361328125
        },
        {
          "model": "joint",
          "length": 8,
          "k": 2,
          "n": 1024,
          "pair_accuracy": 0.0439453125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.04068324609375,
          "target_nll": 11.007974625,
          "top1_on_support": 0.0439453125
        },
        {
          "model": "joint",
          "length": 8,
          "k": 4,
          "n": 1024,
          "pair_accuracy": 0.056640625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0572537060546875,
          "target_nll": 10.2685804375,
          "top1_on_support": 0.056640625
        },
        {
          "model": "joint",
          "length": 8,
          "k": 8,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9570027880859375,
          "target_nll": 3.47798883984375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 8,
          "k": 12,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9570027880859375,
          "target_nll": 3.47798883984375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 8,
          "k": 16,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9570027880859375,
          "target_nll": 3.47798883984375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 8,
          "k": 24,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9570027880859375,
          "target_nll": 3.47798883984375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 12,
          "k": 0,
          "n": 1024,
          "pair_accuracy": 0.0322265625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.032258095703125,
          "target_nll": 20.16554641796875,
          "top1_on_support": 0.0322265625
        },
        {
          "model": "joint",
          "length": 12,
          "k": 1,
          "n": 1024,
          "pair_accuracy": 0.025390625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.029973951171875,
          "target_nll": 11.7212986953125,
          "top1_on_support": 0.025390625
        },
        {
          "model": "joint",
          "length": 12,
          "k": 2,
          "n": 1024,
          "pair_accuracy": 0.0283203125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0305177294921875,
          "target_nll": 11.098264217773437,
          "top1_on_support": 0.0283203125
        },
        {
          "model": "joint",
          "length": 12,
          "k": 4,
          "n": 1024,
          "pair_accuracy": 0.037109375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0395406298828125,
          "target_nll": 10.410232067382813,
          "top1_on_support": 0.037109375
        },
        {
          "model": "joint",
          "length": 12,
          "k": 8,
          "n": 1024,
          "pair_accuracy": 0.0595703125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0555810400390625,
          "target_nll": 9.650599956054688,
          "top1_on_support": 0.0595703125
        },
        {
          "model": "joint",
          "length": 12,
          "k": 12,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9355152841796875,
          "target_nll": 3.50073218359375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 12,
          "k": 16,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9355152841796875,
          "target_nll": 3.50073218359375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 12,
          "k": 24,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9355152841796875,
          "target_nll": 3.50073218359375,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 16,
          "k": 0,
          "n": 1024,
          "pair_accuracy": 0.025390625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0307775009765625,
          "target_nll": 20.191143989257814,
          "top1_on_support": 0.025390625
        },
        {
          "model": "joint",
          "length": 16,
          "k": 1,
          "n": 1024,
          "pair_accuracy": 0.0322265625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0344181005859375,
          "target_nll": 11.673110485351563,
          "top1_on_support": 0.0322265625
        },
        {
          "model": "joint",
          "length": 16,
          "k": 2,
          "n": 1024,
          "pair_accuracy": 0.0234375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.032537095703125,
          "target_nll": 11.084337234375,
          "top1_on_support": 0.0234375
        },
        {
          "model": "joint",
          "length": 16,
          "k": 4,
          "n": 1024,
          "pair_accuracy": 0.03125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0332022548828125,
          "target_nll": 10.465498924804688,
          "top1_on_support": 0.03125
        },
        {
          "model": "joint",
          "length": 16,
          "k": 8,
          "n": 1024,
          "pair_accuracy": 0.0361328125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.037724380859375,
          "target_nll": 9.786500454101562,
          "top1_on_support": 0.0361328125
        },
        {
          "model": "joint",
          "length": 16,
          "k": 12,
          "n": 1024,
          "pair_accuracy": 0.048828125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0519055615234375,
          "target_nll": 9.290152549804688,
          "top1_on_support": 0.048828125
        },
        {
          "model": "joint",
          "length": 16,
          "k": 16,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9158811865234375,
          "target_nll": 3.5220009091796873,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 16,
          "k": 24,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.9158811865234375,
          "target_nll": 3.5220009091796873,
          "top1_on_support": 1.0
        },
        {
          "model": "joint",
          "length": 24,
          "k": 0,
          "n": 1024,
          "pair_accuracy": 0.03515625,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.031281533203125,
          "target_nll": 20.18242931347656,
          "top1_on_support": 0.03515625
        },
        {
          "model": "joint",
          "length": 24,
          "k": 1,
          "n": 1024,
          "pair_accuracy": 0.02734375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.035292025390625,
          "target_nll": 11.676025390625,
          "top1_on_support": 0.02734375
        },
        {
          "model": "joint",
          "length": 24,
          "k": 2,
          "n": 1024,
          "pair_accuracy": 0.0361328125,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.034343091796875,
          "target_nll": 11.06592559765625,
          "top1_on_support": 0.0361328125
        },
        {
          "model": "joint",
          "length": 24,
          "k": 4,
          "n": 1024,
          "pair_accuracy": 0.02734375,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0328869765625,
          "target_nll": 10.450765610351562,
          "top1_on_support": 0.02734375
        },
        {
          "model": "joint",
          "length": 24,
          "k": 8,
          "n": 1024,
          "pair_accuracy": 0.0341796875,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.0321604013671875,
          "target_nll": 9.814048767578125,
          "top1_on_support": 0.0341796875
        },
        {
          "model": "joint",
          "length": 24,
          "k": 12,
          "n": 1024,
          "pair_accuracy": 0.0263671875,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.034471025390625,
          "target_nll": 9.404008388671874,
          "top1_on_support": 0.0263671875
        },
        {
          "model": "joint",
          "length": 24,
          "k": 16,
          "n": 1024,
          "pair_accuracy": 0.0419921875,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.03792431640625,
          "target_nll": 9.102806090820312,
          "top1_on_support": 0.0419921875
        },
        {
          "model": "joint",
          "length": 24,
          "k": 24,
          "n": 1024,
          "pair_accuracy": 1.0,
          "a_accuracy": NaN,
          "b_accuracy": NaN,
          "register_accuracy": NaN,
          "target_mass": 0.8735854326171875,
          "target_nll": 3.5693570380859376,
          "top1_on_support": 1.0
        }
      ]
    }
  ],
  "checkpoints": [
    "large_artifacts/joint_register_executor/checkpoints/main_belief_joint_mod31/checkpoint_step00250.pt",
    "large_artifacts/joint_register_executor/checkpoints/main_belief_joint_mod31/checkpoint_step00500.pt"
  ]
}