PRL: Process Reward Learning Improves LLMs’ Reasoning Ability and Broadens the Reasoning Boundary

This paper introduces Process Reward Learning (PRL), a novel reinforcement learning framework designed to enhance the reasoning capabilities of Large Language Models (LLMs). Unlike traditional methods that rely on sparse "outcome rewards" given only at the end of a task, PRL derives dense, step-by-step supervision signals from a mathematically rigorous decomposition of the global objective. This approach eliminates the need for computationally expensive tools like Monte Carlo Tree Search or separate reward models, significantly boosting training efficiency. Experiments on mathematical benchmarks using models like Qwen2.5-Math and Llama-3.2 show that PRL consistently improves average performance and extends the model's reasoning boundary. Ultimately, the framework provides a theoretical and practical solution for guiding models through complex, multi-step logical challenges.