7 papers
Waypoint-OPD: Transferring RL-Induced Policy Shifts via Gated On-Policy Distillation
Bingxi Hou, Guochao Jiang, Guofeng Quan +4
Policy-shift distillation reuses the difference between a teacher's log probabilities before and after reinforcement learning (RL) as supervision on student-generated trajectories.…
Rethinking Cross-Tokenizer On-Policy Distillation: From Alignment Coverage to Supervision Reliability
Bingxi Hou, Guochao Jiang, Guofeng Quan +4
On-Policy Distillation (OPD) trains a student on its own generations using teacher feedback. With different tokenizers, comparing teacher and student predictions requires alignment…
DVAO: Dynamic Variance-adaptive Advantage Optimization for Multi-reward Reinforcement Learning
Guochao Jiang, Jingyi Song, Guofeng Quan +3
Reinforcement Learning has become a standard paradigm for aligning Large Language Models with human intent and task requirements. While Group Relative Policy Optimization offers an…
Beyond Stochastic Exploration: What Makes Training Data Valuable for Agentic Search
Chuzhan Hao, Wenfeng Feng, Guochao Jiang +3
Reinforcement learning (RL) has become an effective approach for advancing the reasoning capabilities of large language models (LLMs) through the strategic integration of external…
VCRL: Variance-based Curriculum Reinforcement Learning for Large Language Models
Guochao Jiang, Wenfeng Feng, Guofeng Quan +4
Policy-based reinforcement learning currently plays an important role in improving LLMs on mathematical reasoning tasks. However, existing rollout-based reinforcement learning meth…
FlashThink: An Early Exit Method For Efficient Reasoning
Guochao Jiang, Guofeng Quan, Zepeng Ding +3
Large Language Models (LLMs) have shown impressive performance in reasoning tasks. However, LLMs tend to generate excessively long reasoning content, leading to significant computa…
RASD: Retrieval-Augmented Speculative Decoding
Guofeng Quan, Wenfeng Feng, Chuzhan Hao +3
Speculative decoding accelerates inference in large language models (LLMs) by generating draft tokens for target model verification. Current approaches for obtaining draft tokens r…