{
  "mode": "dense",
  "variant": "smoke_gru_mlp_d32",
  "transition": "gru",
  "decoder_type": "mlp",
  "args": {
    "variant_name": "smoke_gru_mlp_d32",
    "mode": "dense",
    "supervision": "teacher_belief",
    "seed": 53,
    "modulus": 7,
    "observe_mod": 3,
    "observe_prob": 0.3,
    "op_family": "full",
    "query_family": "all",
    "train_min_len": 1,
    "train_max_len": 3,
    "eval_lengths": "2,3",
    "eval_k": "0,1,2,3",
    "eval_query_types": "all",
    "state_dim": 32,
    "instr_dim": 32,
    "transition": "gru",
    "decoder_type": "mlp",
    "decoder_rank": 16,
    "heads": 4,
    "compiler_layers": 2,
    "dropout": 0.0,
    "train_steps": 2,
    "batch_size": 16,
    "eval_batch_size": 16,
    "eval_examples": 16,
    "lr": 0.001,
    "weight_decay": 0.0001,
    "query_loss_weight": 0.0,
    "belief_loss_weight": 1.0,
    "decoder_query_loss_weight": 0.0,
    "belief_distill_prob": 1.0,
    "probe_steps": 2,
    "probe_batch_size": 16,
    "probe_lr": 0.001,
    "probe_weight_decay": 0.0001,
    "probe_dropout": 0.0,
    "max_grad_norm": 1.0,
    "log_every": 1,
    "probe_log_every": 1,
    "output_dir": "experiments/dense_teacher_distillation/runs/smoke_gru_mlp_d32",
    "checkpoint_dir": "large_artifacts/dense_teacher_distillation/checkpoints/smoke_gru_mlp_d32"
  },
  "env": {
    "python": "3.12.3",
    "torch": "2.8.0+cu128",
    "device": "cuda",
    "cuda_name": "NVIDIA RTX 6000 Ada Generation"
  },
  "train": [
    {
      "step": 1,
      "loss": 3.9360299110412598,
      "grad_norm": 0.4412754476070404,
      "query_loss": 0.0,
      "belief_loss": 3.9360299110412598,
      "decoder_query_loss": 0.0,
      "query_target_mass": 0.43038204312324524,
      "query_target_nll": 1.9379911422729492,
      "query_top1_on_support": 0.4375,
      "mean_query_support_size": 3.0,
      "decoder_belief_target_mass": 0.09024330973625183,
      "decoder_belief_target_nll": 3.886401653289795,
      "decoder_belief_top1_on_support": 0.125,
      "mean_decoder_belief_support_size": 4.3125,
      "decoder_query_target_mass": 0.4260561168193817,
      "decoder_query_target_nll": 1.9590401649475098,
      "decoder_query_top1_on_support": 0.3125,
      "mean_decoder_query_support_size": 3.0
    },
    {
      "step": 2,
      "loss": 3.8924407958984375,
      "grad_norm": 0.4662870168685913,
      "query_loss": 0.0,
      "belief_loss": 3.8924407958984375,
      "decoder_query_loss": 0.0,
      "query_target_mass": 0.509638249874115,
      "query_target_nll": 1.9547538757324219,
      "query_top1_on_support": 0.5,
      "mean_query_support_size": 3.5625,
      "decoder_belief_target_mass": 0.10067465901374817,
      "decoder_belief_target_nll": 3.848252296447754,
      "decoder_belief_top1_on_support": 0.0625,
      "mean_decoder_belief_support_size": 4.6875,
      "decoder_query_target_mass": 0.5129088759422302,
      "decoder_query_target_nll": 1.933437466621399,
      "decoder_query_top1_on_support": 0.5625,
      "mean_decoder_query_support_size": 3.5625
    }
  ],
  "probe_train": [
    {
      "step": 1,
      "loss": 3.90415620803833,
      "grad_norm": 0.46801742911338806,
      "probe_belief_target_mass": 0.11082718521356583,
      "probe_belief_target_nll": 3.90415620803833,
      "probe_belief_top1_on_support": 0.0625,
      "mean_belief_support_size": 5.359375
    },
    {
      "step": 2,
      "loss": 3.8831682205200195,
      "grad_norm": 0.45102158188819885,
      "probe_belief_target_mass": 0.11645501852035522,
      "probe_belief_target_nll": 3.8831682205200195,
      "probe_belief_top1_on_support": 0.0625,
      "mean_belief_support_size": 5.65625
    }
  ],
  "eval": [
    {
      "step": 2,
      "rows": [
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A",
          "k": 0,
          "n": 16,
          "query_target_mass": 0.6324556469917297,
          "query_target_nll": 1.990732192993164,
          "query_top1_on_support": 0.5,
          "mean_query_support_size": 4.5,
          "probe_belief_target_mass": 0.09367977827787399,
          "probe_belief_target_nll": 3.881014347076416,
          "probe_belief_top1_on_support": 0.0625,
          "mean_belief_support_size": 4.5,
          "decoder_belief_target_mass": 0.09455510973930359,
          "decoder_belief_target_nll": 3.8824071884155273,
          "decoder_belief_top1_on_support": 0.1875,
          "mean_decoder_belief_support_size": 4.5,
          "decoder_query_target_mass": 0.6430114507675171,
          "decoder_query_target_nll": 1.94661283493042,
          "decoder_query_top1_on_support": 0.6875,
          "mean_decoder_query_support_size": 4.5
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A",
          "k": 1,
          "n": 16,
          "query_target_mass": 0.6410203576087952,
          "query_target_nll": 1.9621193408966064,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 4.5,
          "probe_belief_target_mass": 0.09367147088050842,
          "probe_belief_target_nll": 3.8861141204833984,
          "probe_belief_top1_on_support": 0.0625,
          "mean_belief_support_size": 4.5,
          "decoder_belief_target_mass": 0.09257669746875763,
          "decoder_belief_target_nll": 3.904303550720215,
          "decoder_belief_top1_on_support": 0.0625,
          "mean_decoder_belief_support_size": 4.5,
          "decoder_query_target_mass": 0.6416069269180298,
          "decoder_query_target_nll": 1.9518818855285645,
          "decoder_query_top1_on_support": 0.5,
          "mean_decoder_query_support_size": 4.5
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A",
          "k": 2,
          "n": 16,
          "query_target_mass": 0.6389297246932983,
          "query_target_nll": 1.9673993587493896,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 4.5,
          "probe_belief_target_mass": 0.09506183862686157,
          "probe_belief_target_nll": 3.8615593910217285,
          "probe_belief_top1_on_support": 0.0625,
          "mean_belief_support_size": 4.5,
          "decoder_belief_target_mass": 0.09287593513727188,
          "decoder_belief_target_nll": 3.893951892852783,
          "decoder_belief_top1_on_support": 0.1875,
          "mean_decoder_belief_support_size": 4.5,
          "decoder_query_target_mass": 0.6428816914558411,
          "decoder_query_target_nll": 1.9476261138916016,
          "decoder_query_top1_on_support": 0.625,
          "mean_decoder_query_support_size": 4.5
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A",
          "k": 3,
          "n": 16,
          "query_target_mass": 0.6389297246932983,
          "query_target_nll": 1.9673993587493896,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 4.5,
          "probe_belief_target_mass": 0.09506183862686157,
          "probe_belief_target_nll": 3.8615593910217285,
          "probe_belief_top1_on_support": 0.0625,
          "mean_belief_support_size": 4.5,
          "decoder_belief_target_mass": 0.09287593513727188,
          "decoder_belief_target_nll": 3.893951892852783,
          "decoder_belief_top1_on_support": 0.1875,
          "mean_decoder_belief_support_size": 4.5,
          "decoder_query_target_mass": 0.6428816914558411,
          "decoder_query_target_nll": 1.9476261138916016,
          "decoder_query_top1_on_support": 0.625,
          "mean_decoder_query_support_size": 4.5
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "B",
          "k": 0,
          "n": 16,
          "query_target_mass": 0.6084702610969543,
          "query_target_nll": 1.9502887725830078,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 4.25,
          "probe_belief_target_mass": 0.09692889451980591,
          "probe_belief_target_nll": 3.914336919784546,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 4.6875,
          "decoder_belief_target_mass": 0.09859812259674072,
          "decoder_belief_target_nll": 3.887972831726074,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 4.6875,
          "decoder_query_target_mass": 0.605815052986145,
          "decoder_query_target_nll": 1.952412486076355,
          "decoder_query_top1_on_support": 0.5625,
          "mean_decoder_query_support_size": 4.25
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "B",
          "k": 1,
          "n": 16,
          "query_target_mass": 0.6130075454711914,
          "query_target_nll": 1.9282759428024292,
          "query_top1_on_support": 0.625,
          "mean_query_support_size": 4.25,
          "probe_belief_target_mass": 0.09696383029222488,
          "probe_belief_target_nll": 3.906874179840088,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 4.6875,
          "decoder_belief_target_mass": 0.0982150137424469,
          "decoder_belief_target_nll": 3.8895559310913086,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 4.6875,
          "decoder_query_target_mass": 0.6089703440666199,
          "decoder_query_target_nll": 1.9468307495117188,
          "decoder_query_top1_on_support": 0.625,
          "mean_decoder_query_support_size": 4.25
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "B",
          "k": 2,
          "n": 16,
          "query_target_mass": 0.610136091709137,
          "query_target_nll": 1.9405243396759033,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 4.25,
          "probe_belief_target_mass": 0.09677217155694962,
          "probe_belief_target_nll": 3.9103684425354004,
          "probe_belief_top1_on_support": 0.3125,
          "mean_belief_support_size": 4.6875,
          "decoder_belief_target_mass": 0.09822480380535126,
          "decoder_belief_target_nll": 3.886357307434082,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 4.6875,
          "decoder_query_target_mass": 0.6093337535858154,
          "decoder_query_target_nll": 1.9401686191558838,
          "decoder_query_top1_on_support": 0.625,
          "mean_decoder_query_support_size": 4.25
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "B",
          "k": 3,
          "n": 16,
          "query_target_mass": 0.610136091709137,
          "query_target_nll": 1.9405243396759033,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 4.25,
          "probe_belief_target_mass": 0.09677217155694962,
          "probe_belief_target_nll": 3.9103684425354004,
          "probe_belief_top1_on_support": 0.3125,
          "mean_belief_support_size": 4.6875,
          "decoder_belief_target_mass": 0.09822480380535126,
          "decoder_belief_target_nll": 3.886357307434082,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 4.6875,
          "decoder_query_target_mass": 0.6093337535858154,
          "decoder_query_target_nll": 1.9401684999465942,
          "decoder_query_top1_on_support": 0.625,
          "mean_decoder_query_support_size": 4.25
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A_PLUS_B",
          "k": 0,
          "n": 16,
          "query_target_mass": 0.6065726280212402,
          "query_target_nll": 1.9876105785369873,
          "query_top1_on_support": 0.4375,
          "mean_query_support_size": 4.3125,
          "probe_belief_target_mass": 0.08821803331375122,
          "probe_belief_target_nll": 3.8992743492126465,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 4.3125,
          "decoder_belief_target_mass": 0.09040525555610657,
          "decoder_belief_target_nll": 3.8696231842041016,
          "decoder_belief_top1_on_support": 0.0625,
          "mean_decoder_belief_support_size": 4.3125,
          "decoder_query_target_mass": 0.6181036829948425,
          "decoder_query_target_nll": 1.939653992652893,
          "decoder_query_top1_on_support": 0.75,
          "mean_decoder_query_support_size": 4.3125
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A_PLUS_B",
          "k": 1,
          "n": 16,
          "query_target_mass": 0.6057557463645935,
          "query_target_nll": 1.9998250007629395,
          "query_top1_on_support": 0.4375,
          "mean_query_support_size": 4.3125,
          "probe_belief_target_mass": 0.0871468335390091,
          "probe_belief_target_nll": 3.9171953201293945,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 4.3125,
          "decoder_belief_target_mass": 0.08993060886859894,
          "decoder_belief_target_nll": 3.8970375061035156,
          "decoder_belief_top1_on_support": 0.0625,
          "mean_decoder_belief_support_size": 4.3125,
          "decoder_query_target_mass": 0.6147352457046509,
          "decoder_query_target_nll": 1.9516024589538574,
          "decoder_query_top1_on_support": 0.5625,
          "mean_decoder_query_support_size": 4.3125
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A_PLUS_B",
          "k": 2,
          "n": 16,
          "query_target_mass": 0.6067408323287964,
          "query_target_nll": 2.0027170181274414,
          "query_top1_on_support": 0.5,
          "mean_query_support_size": 4.3125,
          "probe_belief_target_mass": 0.08518335223197937,
          "probe_belief_target_nll": 3.9415814876556396,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 4.3125,
          "decoder_belief_target_mass": 0.08979468047618866,
          "decoder_belief_target_nll": 3.899989128112793,
          "decoder_belief_top1_on_support": 0.125,
          "mean_decoder_belief_support_size": 4.3125,
          "decoder_query_target_mass": 0.6148179769515991,
          "decoder_query_target_nll": 1.9511635303497314,
          "decoder_query_top1_on_support": 0.4375,
          "mean_decoder_query_support_size": 4.3125
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A_PLUS_B",
          "k": 3,
          "n": 16,
          "query_target_mass": 0.6067408323287964,
          "query_target_nll": 2.0027170181274414,
          "query_top1_on_support": 0.5,
          "mean_query_support_size": 4.3125,
          "probe_belief_target_mass": 0.08518335223197937,
          "probe_belief_target_nll": 3.9415814876556396,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 4.3125,
          "decoder_belief_target_mass": 0.08979468047618866,
          "decoder_belief_target_nll": 3.899989128112793,
          "decoder_belief_top1_on_support": 0.125,
          "mean_decoder_belief_support_size": 4.3125,
          "decoder_query_target_mass": 0.6148179769515991,
          "decoder_query_target_nll": 1.9511635303497314,
          "decoder_query_top1_on_support": 0.4375,
          "mean_decoder_query_support_size": 4.3125
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A_MINUS_B",
          "k": 0,
          "n": 16,
          "query_target_mass": 0.4550463557243347,
          "query_target_nll": 2.020543098449707,
          "query_top1_on_support": 0.4375,
          "mean_query_support_size": 3.25,
          "probe_belief_target_mass": 0.08796326816082001,
          "probe_belief_target_nll": 3.9206552505493164,
          "probe_belief_top1_on_support": 0.25,
          "mean_belief_support_size": 4.25,
          "decoder_belief_target_mass": 0.08545739203691483,
          "decoder_belief_target_nll": 3.9136197566986084,
          "decoder_belief_top1_on_support": 0.125,
          "mean_decoder_belief_support_size": 4.25,
          "decoder_query_target_mass": 0.46137532591819763,
          "decoder_query_target_nll": 1.9657917022705078,
          "decoder_query_top1_on_support": 0.4375,
          "mean_decoder_query_support_size": 3.25
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A_MINUS_B",
          "k": 1,
          "n": 16,
          "query_target_mass": 0.4515995681285858,
          "query_target_nll": 2.0381526947021484,
          "query_top1_on_support": 0.5,
          "mean_query_support_size": 3.25,
          "probe_belief_target_mass": 0.08952361345291138,
          "probe_belief_target_nll": 3.8991293907165527,
          "probe_belief_top1_on_support": 0.1875,
          "mean_belief_support_size": 4.25,
          "decoder_belief_target_mass": 0.08575551211833954,
          "decoder_belief_target_nll": 3.910299301147461,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 4.25,
          "decoder_query_target_mass": 0.46436309814453125,
          "decoder_query_target_nll": 1.945643663406372,
          "decoder_query_top1_on_support": 0.4375,
          "mean_decoder_query_support_size": 3.25
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A_MINUS_B",
          "k": 2,
          "n": 16,
          "query_target_mass": 0.45199573040008545,
          "query_target_nll": 2.0322043895721436,
          "query_top1_on_support": 0.4375,
          "mean_query_support_size": 3.25,
          "probe_belief_target_mass": 0.08960415422916412,
          "probe_belief_target_nll": 3.904116153717041,
          "probe_belief_top1_on_support": 0.1875,
          "mean_belief_support_size": 4.25,
          "decoder_belief_target_mass": 0.0866924524307251,
          "decoder_belief_target_nll": 3.919755458831787,
          "decoder_belief_top1_on_support": 0.0625,
          "mean_decoder_belief_support_size": 4.25,
          "decoder_query_target_mass": 0.46616438031196594,
          "decoder_query_target_nll": 1.937260389328003,
          "decoder_query_top1_on_support": 0.625,
          "mean_decoder_query_support_size": 3.25
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 2,
          "query_type": "A_MINUS_B",
          "k": 3,
          "n": 16,
          "query_target_mass": 0.45199573040008545,
          "query_target_nll": 2.0322043895721436,
          "query_top1_on_support": 0.4375,
          "mean_query_support_size": 3.25,
          "probe_belief_target_mass": 0.08960415422916412,
          "probe_belief_target_nll": 3.904116153717041,
          "probe_belief_top1_on_support": 0.1875,
          "mean_belief_support_size": 4.25,
          "decoder_belief_target_mass": 0.0866924524307251,
          "decoder_belief_target_nll": 3.919755458831787,
          "decoder_belief_top1_on_support": 0.0625,
          "mean_decoder_belief_support_size": 4.25,
          "decoder_query_target_mass": 0.46616441011428833,
          "decoder_query_target_nll": 1.937260389328003,
          "decoder_query_top1_on_support": 0.625,
          "mean_decoder_query_support_size": 3.25
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A",
          "k": 0,
          "n": 16,
          "query_target_mass": 0.4098360538482666,
          "query_target_nll": 1.9515299797058105,
          "query_top1_on_support": 0.375,
          "mean_query_support_size": 2.875,
          "probe_belief_target_mass": 0.06683404743671417,
          "probe_belief_target_nll": 3.901224136352539,
          "probe_belief_top1_on_support": 0.0,
          "mean_belief_support_size": 3.3125,
          "decoder_belief_target_mass": 0.0681302547454834,
          "decoder_belief_target_nll": 3.9380178451538086,
          "decoder_belief_top1_on_support": 0.125,
          "mean_decoder_belief_support_size": 3.3125,
          "decoder_query_target_mass": 0.4095776081085205,
          "decoder_query_target_nll": 1.9552102088928223,
          "decoder_query_top1_on_support": 0.375,
          "mean_decoder_query_support_size": 2.875
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A",
          "k": 1,
          "n": 16,
          "query_target_mass": 0.4091036915779114,
          "query_target_nll": 1.9680242538452148,
          "query_top1_on_support": 0.375,
          "mean_query_support_size": 2.875,
          "probe_belief_target_mass": 0.06694436073303223,
          "probe_belief_target_nll": 3.8946266174316406,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 3.3125,
          "decoder_belief_target_mass": 0.06767958402633667,
          "decoder_belief_target_nll": 3.9466724395751953,
          "decoder_belief_top1_on_support": 0.0625,
          "mean_decoder_belief_support_size": 3.3125,
          "decoder_query_target_mass": 0.40922969579696655,
          "decoder_query_target_nll": 1.9565948247909546,
          "decoder_query_top1_on_support": 0.375,
          "mean_decoder_query_support_size": 2.875
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A",
          "k": 2,
          "n": 16,
          "query_target_mass": 0.41119006276130676,
          "query_target_nll": 1.9575018882751465,
          "query_top1_on_support": 0.4375,
          "mean_query_support_size": 2.875,
          "probe_belief_target_mass": 0.06711472570896149,
          "probe_belief_target_nll": 3.891462802886963,
          "probe_belief_top1_on_support": 0.1875,
          "mean_belief_support_size": 3.3125,
          "decoder_belief_target_mass": 0.06865906715393066,
          "decoder_belief_target_nll": 3.9242539405822754,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 3.3125,
          "decoder_query_target_mass": 0.41161954402923584,
          "decoder_query_target_nll": 1.947417140007019,
          "decoder_query_top1_on_support": 0.5,
          "mean_decoder_query_support_size": 2.875
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A",
          "k": 3,
          "n": 16,
          "query_target_mass": 0.40994226932525635,
          "query_target_nll": 1.9623534679412842,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 2.875,
          "probe_belief_target_mass": 0.06642170995473862,
          "probe_belief_target_nll": 3.9079818725585938,
          "probe_belief_top1_on_support": 0.1875,
          "mean_belief_support_size": 3.3125,
          "decoder_belief_target_mass": 0.06863594055175781,
          "decoder_belief_target_nll": 3.925351619720459,
          "decoder_belief_top1_on_support": 0.0625,
          "mean_decoder_belief_support_size": 3.3125,
          "decoder_query_target_mass": 0.40817657113075256,
          "decoder_query_target_nll": 1.9613397121429443,
          "decoder_query_top1_on_support": 0.3125,
          "mean_decoder_query_support_size": 2.875
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "B",
          "k": 0,
          "n": 16,
          "query_target_mass": 0.3985596001148224,
          "query_target_nll": 1.9550877809524536,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 2.75,
          "probe_belief_target_mass": 0.08084684610366821,
          "probe_belief_target_nll": 3.879804849624634,
          "probe_belief_top1_on_support": 0.25,
          "mean_belief_support_size": 3.8125,
          "decoder_belief_target_mass": 0.07889526337385178,
          "decoder_belief_target_nll": 3.900233268737793,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 3.8125,
          "decoder_query_target_mass": 0.3934136927127838,
          "decoder_query_target_nll": 1.9413299560546875,
          "decoder_query_top1_on_support": 0.25,
          "mean_decoder_query_support_size": 2.75
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "B",
          "k": 1,
          "n": 16,
          "query_target_mass": 0.40413862466812134,
          "query_target_nll": 1.9351544380187988,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 2.75,
          "probe_belief_target_mass": 0.07757723331451416,
          "probe_belief_target_nll": 3.911406993865967,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 3.8125,
          "decoder_belief_target_mass": 0.07717521488666534,
          "decoder_belief_target_nll": 3.9078927040100098,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 3.8125,
          "decoder_query_target_mass": 0.3894929885864258,
          "decoder_query_target_nll": 1.9582529067993164,
          "decoder_query_top1_on_support": 0.1875,
          "mean_decoder_query_support_size": 2.75
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "B",
          "k": 2,
          "n": 16,
          "query_target_mass": 0.40474170446395874,
          "query_target_nll": 1.912348985671997,
          "query_top1_on_support": 0.625,
          "mean_query_support_size": 2.75,
          "probe_belief_target_mass": 0.07790543138980865,
          "probe_belief_target_nll": 3.9222490787506104,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 3.8125,
          "decoder_belief_target_mass": 0.07982347905635834,
          "decoder_belief_target_nll": 3.873781442642212,
          "decoder_belief_top1_on_support": 0.125,
          "mean_decoder_belief_support_size": 3.8125,
          "decoder_query_target_mass": 0.3908456563949585,
          "decoder_query_target_nll": 1.9541733264923096,
          "decoder_query_top1_on_support": 0.25,
          "mean_decoder_query_support_size": 2.75
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "B",
          "k": 3,
          "n": 16,
          "query_target_mass": 0.40148282051086426,
          "query_target_nll": 1.9197852611541748,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 2.75,
          "probe_belief_target_mass": 0.07970146834850311,
          "probe_belief_target_nll": 3.8962061405181885,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 3.8125,
          "decoder_belief_target_mass": 0.08000479638576508,
          "decoder_belief_target_nll": 3.8941214084625244,
          "decoder_belief_top1_on_support": 0.125,
          "mean_decoder_belief_support_size": 3.8125,
          "decoder_query_target_mass": 0.38898980617523193,
          "decoder_query_target_nll": 1.963921070098877,
          "decoder_query_top1_on_support": 0.375,
          "mean_decoder_query_support_size": 2.75
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A_PLUS_B",
          "k": 0,
          "n": 16,
          "query_target_mass": 0.4622237980365753,
          "query_target_nll": 1.948227047920227,
          "query_top1_on_support": 0.5,
          "mean_query_support_size": 3.1875,
          "probe_belief_target_mass": 0.06601977348327637,
          "probe_belief_target_nll": 3.889474630355835,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 3.1875,
          "decoder_belief_target_mass": 0.06437428295612335,
          "decoder_belief_target_nll": 3.932417392730713,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 3.1875,
          "decoder_query_target_mass": 0.4550204575061798,
          "decoder_query_target_nll": 1.9512090682983398,
          "decoder_query_top1_on_support": 0.4375,
          "mean_decoder_query_support_size": 3.1875
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A_PLUS_B",
          "k": 1,
          "n": 16,
          "query_target_mass": 0.45794135332107544,
          "query_target_nll": 1.9605004787445068,
          "query_top1_on_support": 0.5,
          "mean_query_support_size": 3.1875,
          "probe_belief_target_mass": 0.06476031243801117,
          "probe_belief_target_nll": 3.8985116481781006,
          "probe_belief_top1_on_support": 0.0625,
          "mean_belief_support_size": 3.1875,
          "decoder_belief_target_mass": 0.06308506429195404,
          "decoder_belief_target_nll": 3.9546704292297363,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 3.1875,
          "decoder_query_target_mass": 0.45419836044311523,
          "decoder_query_target_nll": 1.9530839920043945,
          "decoder_query_top1_on_support": 0.375,
          "mean_decoder_query_support_size": 3.1875
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A_PLUS_B",
          "k": 2,
          "n": 16,
          "query_target_mass": 0.45407429337501526,
          "query_target_nll": 1.9778820276260376,
          "query_top1_on_support": 0.4375,
          "mean_query_support_size": 3.1875,
          "probe_belief_target_mass": 0.0663640946149826,
          "probe_belief_target_nll": 3.875100612640381,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 3.1875,
          "decoder_belief_target_mass": 0.06348589807748795,
          "decoder_belief_target_nll": 3.940614700317383,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 3.1875,
          "decoder_query_target_mass": 0.4565083980560303,
          "decoder_query_target_nll": 1.941530466079712,
          "decoder_query_top1_on_support": 0.4375,
          "mean_decoder_query_support_size": 3.1875
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A_PLUS_B",
          "k": 3,
          "n": 16,
          "query_target_mass": 0.454275518655777,
          "query_target_nll": 1.9834778308868408,
          "query_top1_on_support": 0.5,
          "mean_query_support_size": 3.1875,
          "probe_belief_target_mass": 0.06612588465213776,
          "probe_belief_target_nll": 3.8863606452941895,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 3.1875,
          "decoder_belief_target_mass": 0.06517767906188965,
          "decoder_belief_target_nll": 3.9110569953918457,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 3.1875,
          "decoder_query_target_mass": 0.4566951096057892,
          "decoder_query_target_nll": 1.9385567903518677,
          "decoder_query_top1_on_support": 0.5,
          "mean_decoder_query_support_size": 3.1875
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A_MINUS_B",
          "k": 0,
          "n": 16,
          "query_target_mass": 0.5418770909309387,
          "query_target_nll": 1.9600080251693726,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 3.8125,
          "probe_belief_target_mass": 0.09164488315582275,
          "probe_belief_target_nll": 3.975338935852051,
          "probe_belief_top1_on_support": 0.0,
          "mean_belief_support_size": 4.625,
          "decoder_belief_target_mass": 0.0935862809419632,
          "decoder_belief_target_nll": 3.919703483581543,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 4.625,
          "decoder_query_target_mass": 0.5465324521064758,
          "decoder_query_target_nll": 1.9443809986114502,
          "decoder_query_top1_on_support": 0.5,
          "mean_decoder_query_support_size": 3.8125
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A_MINUS_B",
          "k": 1,
          "n": 16,
          "query_target_mass": 0.54667067527771,
          "query_target_nll": 1.9454764127731323,
          "query_top1_on_support": 0.5625,
          "mean_query_support_size": 3.8125,
          "probe_belief_target_mass": 0.09266746044158936,
          "probe_belief_target_nll": 3.9591164588928223,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 4.625,
          "decoder_belief_target_mass": 0.09451449662446976,
          "decoder_belief_target_nll": 3.915893077850342,
          "decoder_belief_top1_on_support": 0.0625,
          "mean_decoder_belief_support_size": 4.625,
          "decoder_query_target_mass": 0.5454455018043518,
          "decoder_query_target_nll": 1.9475290775299072,
          "decoder_query_top1_on_support": 0.5,
          "mean_decoder_query_support_size": 3.8125
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A_MINUS_B",
          "k": 2,
          "n": 16,
          "query_target_mass": 0.5456092953681946,
          "query_target_nll": 1.94820237159729,
          "query_top1_on_support": 0.625,
          "mean_query_support_size": 3.8125,
          "probe_belief_target_mass": 0.0932612493634224,
          "probe_belief_target_nll": 3.9468634128570557,
          "probe_belief_top1_on_support": 0.0,
          "mean_belief_support_size": 4.625,
          "decoder_belief_target_mass": 0.09373864531517029,
          "decoder_belief_target_nll": 3.9278268814086914,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 4.625,
          "decoder_query_target_mass": 0.5463315844535828,
          "decoder_query_target_nll": 1.9423243999481201,
          "decoder_query_top1_on_support": 0.5625,
          "mean_decoder_query_support_size": 3.8125
        },
        {
          "model": "dense",
          "variant": "smoke_gru_mlp_d32",
          "supervision": "teacher_belief",
          "transition": "gru",
          "decoder_type": "mlp",
          "state_dim": 32,
          "instr_dim": 32,
          "decoder_rank": 16,
          "length": 3,
          "query_type": "A_MINUS_B",
          "k": 3,
          "n": 16,
          "query_target_mass": 0.5489360690116882,
          "query_target_nll": 1.9433895349502563,
          "query_top1_on_support": 0.625,
          "mean_query_support_size": 3.8125,
          "probe_belief_target_mass": 0.09404687583446503,
          "probe_belief_target_nll": 3.938730239868164,
          "probe_belief_top1_on_support": 0.125,
          "mean_belief_support_size": 4.625,
          "decoder_belief_target_mass": 0.09439843147993088,
          "decoder_belief_target_nll": 3.9183712005615234,
          "decoder_belief_top1_on_support": 0.0,
          "mean_decoder_belief_support_size": 4.625,
          "decoder_query_target_mass": 0.5470160841941833,
          "decoder_query_target_nll": 1.9411098957061768,
          "decoder_query_top1_on_support": 0.75,
          "mean_decoder_query_support_size": 3.8125
        }
      ]
    }
  ],
  "checkpoints": [
    "large_artifacts/dense_teacher_distillation/checkpoints/smoke_gru_mlp_d32/checkpoint_final.pt"
  ]
}