Improving Physics Reasoning in Large Language Models Using Mixture of Refinement Agents
This addresses the problem of poor physics reasoning in open-source LLMs for researchers and developers, but it is incremental as it builds on existing refinement and agent-based methods.
The paper tackles the challenge of improving physics reasoning in large language models by introducing Mixture of Refinement Agents (MoRA), which iteratively refines solutions to correct errors, resulting in up to a 16% increase in accuracy for models like Llama-3-70B and Gemma-2-27B on datasets including SciEval, MMLU, and PhysicsQA.
Large Language Models (LLMs) demonstrate remarkable capabilities in various reasoning tasks. However, they encounter significant challenges when it comes to scientific reasoning, particularly in physics, which requires not only mathematical reasoning but also factual and conceptual understanding. When addressing complex physics problems, LLMs typically face three key issues: problem miscomprehension, incorrect concept application, and computational errors. While each of these problems can be addressed individually, there is a need for a generalized approach that can tackle all three issues simultaneously. To address this, we introduce Mixture of Refinement Agents (MoRA), a novel agentic refinement framework that iteratively refines the LLM generated base solution by correcting the aforementioned errors, resulting in a significant performance improvement for open-source LLMs. Our approach aims to bridge the gap between opensource LLMs and GPT-4o by utilizing the latter as error identifier to guide these refinement agents. We evaluate our approach on the SciEval and MMLU subsets along with our own physics dataset (PhysicsQA). MoRA significantly improves the performance of Llama-3-70B and Gemma-2-27B on these datasets, achieving up to a 16% increase in final answer accuracy.