artifacts/baselines/random_baseline.json

{
  "backend": "Random baseline (uniform over valid verbs)",
  "model": "random",
  "env": "https://akhiilll-claims-env.hf.space",
  "episodes": 20,
  "max_steps": 8,
  "rewards": [
    10.975,
    -8.2,
    13.5,
    -1.5000000000000002,
    13.449152542372882,
    12.1,
    15.9,
    -7.1,
    -15.8,
    -7.8,
    16.0,
    13.7,
    -7.3,
    -5.0,
    10.3,
    16.0,
    16.0,
    -15.0,
    11.375,
    11.0
  ],
  "averages": [
    10.975,
    1.3875000000000002,
    5.425,
    3.69375,
    5.644830508474577,
    5.8698305084745765,
    10.689830508474577,
    6.569830508474577,
    3.7098305084745764,
    -0.54,
    0.24000000000000005,
    -0.20000000000000018,
    -0.2400000000000002,
    1.92,
    5.54,
    5.54,
    6.0,
    4.46,
    7.735,
    7.875
  ],
  "start_avg": 10.975,
  "final_avg": 7.875
}