SRPO: A Cross-Domain Implementation of Large-Scale Reinforcement Learning on LLM
By: Xiaojiang Zhang , Jinghui Wang , Zifei Cheng and more
Potential Business Impact:
Makes AI smarter at math and coding faster.
Recent advances of reasoning models, exemplified by OpenAI's o1 and DeepSeek's R1, highlight the significant potential of Reinforcement Learning (RL) to enhance the reasoning capabilities of Large Language Models (LLMs). However, replicating these advancements across diverse domains remains challenging due to limited methodological transparency. In this work, we present two-Staged history-Resampling Policy Optimization (SRPO), which surpasses the performance of DeepSeek-R1-Zero-32B on the AIME24 and LiveCodeBench benchmarks. SRPO achieves this using the same base model as DeepSeek (i.e. Qwen2.5-32B), using only about 1/10 of the training steps required by DeepSeek-R1-Zero-32B, demonstrating superior efficiency. Building upon Group Relative Policy Optimization (GRPO), we introduce two key methodological innovations: (1) a two-stage cross-domain training paradigm designed to balance the development of mathematical reasoning and coding proficiency, and (2) History Resampling (HR), a technique to address ineffective samples. Our comprehensive experiments validate the effectiveness of our approach, offering valuable insights into scaling LLM reasoning capabilities across diverse tasks.
Similar Papers
SRPO: Enhancing Multimodal LLM Reasoning via Reflection-Aware Reinforcement Learning
Computation and Language
Helps AI think better and fix its own mistakes.
SRPO: Self-Referential Policy Optimization for Vision-Language-Action Models
Robotics
Teaches robots to learn from their own mistakes.
RSPO: Regularized Self-Play Alignment of Large Language Models
Machine Learning (CS)
Makes AI assistants more helpful and less wordy.