paper_type.json (291B)
1 { 2 "paper_type": "empirical", 3 "reason": "Proposes a reinforcement learning framework (CWRPO) and validates it experimentally across 12 benchmarks, reporting quantitative improvements over multiple baselines—the primary contribution is experimental findings, not the benchmark itself." 4 }