{
  "title": "Reusing Trajectories in Policy Gradients Enables Fast Convergence",
  "orid": "gcYvvxTLRA",
  "space_id": "neonforestmist/rtpg-trajectory-reuse-repro",
  "forecast": "8/8",
  "repair": "claim-faithful-v3",
  "repaired_at": "2026-07-27T19:01:43.130307+00:00",
  "pages": [
    "01-introduces-rpg-retrospective-reusing-trajectorie",
    "02-derives-concentration-bound-power-mean-estimator",
    "03-rpg-attains-sample-complexity-reach",
    "04-compares-against-reinforce-svrpg-srvrpg",
    "conclusion"
  ],
  "artifacts": [
    "evidence/claim_1.json",
    "evidence/claim_2.json",
    "evidence/claim_3.json",
    "evidence/claim_4.json"
  ],
  "cpu_only": true
}