{
  "scene": "illusion_0.75L",
  "n_steps": 500,
  "observables": [
    "force_total",
    "sensor_error",
    "sensor_error_delayed",
    "p_sig_stack",
    "actions"
  ]
}