Squeeze the Soaked Sponge: Efficient Off-policy Reinforcement Finetuning for Large Language Model

Jul 9, 2025·

Jing Liang

,

Hongyao Tang

,

Yi Ma

Jinyi Liu

,

Yan Zheng

,

Shuyue Hu

,

Lei Bai

,

Jianye Hao

· 0 min read

PDF Cite Source Document

LLM Post-Training (RL Tuning) LLM Post-Training DRL Fine-Tuning Off-Policy

Authors

Jinyi Liu

Ph.D. Candidate

---

← RoboAnnotatorX: A Comprehensive and Universal Annotation Framework for Accurate Understanding of Long-horizon Robot Demonstration Jul 10, 2025

Unlocking Multi-Agent Debate Potential: Enhancing Effective Scaling through Role Allocation Strategies Jun 15, 2025 →