AWhat changed
Combined asymmetric probability-ratio clipping, filtering of zero-advantage prompt groups and token-weighted policy optimization in an openly documented large-scale reasoning-model RL pipeline.
Scientific scope
Combined asymmetric probability-ratio clipping, filtering of zero-advantage prompt groups and token-weighted policy optimization in an openly documented large-scale reasoning-model RL pipeline. PPO, GRPO and verl are inherited. The filtering changes the sampled prompt distribution; source-reported benchmark gains do not prove universal superiority or general safety. Thirty-five unique authors are taken from the explicitly designated Contributions author list, excluding acknowledgments.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. PPO, GRPO and verl are inherited. The filtering changes the sampled prompt distribution; source-reported benchmark gains do not prove universal superiority or general safety. Thirty-five unique authors are taken from the explicitly designated Contributions author list, excluding acknowledgments.
Contributors
- Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
- DAPO: An Open-Source LLM Reinforcement Learning System at Scale
- Yu et al.: DAPO decoupled clipping and dynamic-sampling RL system — author/institution implementation
Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction