← authors · overview

• Defeating the Training-Inference Mismatch via FP16
  2025 · 34 citations
• EnvPool: A Highly Parallel Reinforcement Learning Environment Execution Engine
  2022 · 14 citations
• Riemannian Denoising Diffusion Probabilistic Models
  2025 · 9 citations
• Part II: ROLL Flash -- Accelerating RLVR and Agentic Training with Asynchrony
  2025 · 5 citations
• Part II: ROLL Flash -- Accelerating RLVR and Agentic Training with Asynchrony
  2025 · 5 citations
• Open RL Benchmark: Comprehensive Tracked Experiments for Reinforcement Learning
  2024 · 3 citations
• Reinforcement Learning Optimization for Large-Scale Learning: An Efficient and User-Friendly Scaling Library
  2025 · 1 citations
• MobileGym: A Verifiable and Highly Parallel Simulation Platform for Mobile GUI Agent Research
  2026
• RetroAgent: From Solving to Evolving via Retrospective Dual Intrinsic Feedback
  2026
• RollArt: Scaling Agentic RL Training via Disaggregated Infrastructure
  2025
• GRID: Graph Representation of Intelligence Data for Security Text Knowledge Graph Construction
  2026
• From $P(y|x)$ to $P(y)$: Investigating Reinforcement Learning in Pre-train Space
  2026
• Rethinking the Trust Region in LLM Reinforcement Learning
  2026
• GEM: A Gym for Agentic LLMs
  2025
• Language Models Can Learn from Verbal Feedback Without Scalar Rewards
  2025