DGRO: Enhancing LLM Reasoning via Exploration-Exploitation Control and Reward Variance Management

TL;DR

This paper introduces DGRO, a novel RL algorithm for LLM reasoning that decouples exploration and exploitation controls and manages reward variance, leading to improved reasoning performance and faster convergence.

Contribution

DGRO is a new RL method that independently tunes exploration and exploitation, and incorporates reward variance management, enhancing LLM reasoning capabilities.

Findings

Achieves 96.9% accuracy on Logic dataset

Outperforms existing methods on mathematical benchmarks

Demonstrates improved convergence speed and generalization

Abstract

Inference scaling further accelerates Large Language Models (LLMs) toward Artificial General Intelligence (AGI), with large-scale Reinforcement Learning (RL) to unleash long Chain-of-Thought reasoning. Most contemporary reasoning approaches usually rely on handcrafted rule-based reward functions. However, the tarde-offs of exploration and exploitation in RL algorithms involves multiple complex considerations, and the theoretical and empirical impacts of manually designed reward functions remain insufficiently explored. In this paper, we propose Decoupled Group Reward Optimization (DGRO), a general RL algorithm for LLM reasoning. On the one hand, DGRO decouples the traditional regularization coefficient into two independent hyperparameters: one scales the policy gradient term, and the other regulates the distance from the sampling policy. This decoupling not only enables precise control over balancing exploration and exploitation, but also can be seamlessly extended to Online Policy Mirror Descent (OPMD) algorithms in Kimi k1.5 and Direct Reward Optimization. On the other hand, we observe that reward variance significantly affects both convergence speed and final model performance. We conduct both theoretical analysis and extensive empirical validation to assess DGRO, including a detailed ablation study that investigates its performance and optimization dynamics. Experimental results show that DGRO achieves state-of-the-art performance on the Logic dataset with an average accuracy of 96.9\%, and demonstrates strong generalization across mathematical benchmarks.