LM2: A Simple Society of Language Models Solves Complex Reasoning
Gurusha Juneja, Subhabrata Dutta, Tanmoy Chakraborty
摘要
Despite demonstrating emergent reasoning abilities, Large Language Models (LLMS) often lose track of complex, multi-step reasoning. Existing studies show that providing guidance via decomposing the original question into multiple subproblems elicits more robustness in LLM reasoning -a decomposer generates the subproblems, and a solver solves each of these subproblems. However, these techniques fail to accommodate coordination between the decomposer and the solver modules (either in a single model or different specialized ones) -the decomposer does not keep track of the ability of the solver to follow the decomposed reasoning. In this paper, we propose LM 2 to address these challenges. LM 2 modularizes the decomposition, solution, and verification into three different language models. The decomposer module identifies the key concepts necessary to solve the problem and generates step-by-step subquestions according to the reasoning requirement. The solver model generates the solution to the subproblems that are then checked by the verifier module; depending upon the feedback from the verifier, the reasoning context is constructed using the subproblems and the solutions. These models are trained to coordinate using policy learning. Exhaustive experimentation suggests the superiority of LM 2 over existing methods on in-and out-domain reasoning problems, outperforming the best baselines by 8.1% on MATH, 7.71% on JEEBench, and 9.7% on MedQA problems (code available at https://github.com/ LCS2-IIITD/Language_Model_Multiplex ). : How many distinct, non-equilateral triangles with a perimeter of 60 units have integer side lengths , , and such that , , is an arithmetic sequence? Solver LM Decomposer LM Verifier LM SQ: What is a, b, c in terms of common difference d? SA:Since a, b, and c form an arithmetic sequence, we can express c in terms of a as c = a + d, where d is the common difference Let be the common difference, so and We can assume that is positive In particular, can't be 0, because the triangle is not equilateral Then the perimeter of the triangle is , so Hence, the sides of the triangle are , 20, and These sides must satisfy the triangle inequality, which gives us Solving for , we find , or Therefore, the possible values of are 1, 2, , 9, which gives us possible triangles
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了每一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。
引用它的顶会 Paper3
- Emergent Coordination in Multi-Agent Language ModelsChristoph RiedlICLR 2026 · 被引用 25 次
- RoboAgent: Chaining Basic Capabilities for Embodied Task PlanningPeiran Xu, Jiaqi Zheng, Yadong MuCVPR 2026 · 被引用 6 次
- PaT: Planning-after-Trial for Efficient Test-Time Code GenerationYoungsik Yoon, Sungjae Lee, Seockbean Song, Siwei Wang 等ACL 2026
它引用的顶会 Paper9
- Chain-of-Thought Prompting Elicits Reasoning in Large Language ModelsJason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma 等NeurIPS 2022 · 被引用 22,562 次
- LoRA: Low-Rank Adaptation of Large Language ModelsEdward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu 等ICLR 2022 · 被引用 18,833 次
- Tree of Thoughts: Deliberate Problem Solving with Large Language ModelsShunyu Yao, Dian Yu, Jeffrey Zhao, Izhak Shafran 等NeurIPS 2023 · 被引用 5,068 次
- Least-to-Most Prompting Enables Complex Reasoning in Large Language ModelsDenny Zhou, Nathanael Schärli, Le Hou, Jason Wei 等ICLR 2023 · 被引用 318 次
- Decomposed Prompting: A Modular Approach for Solving Complex TasksTushar Khot, Harsh Trivedi, Matthew Finlayson, Yao Fu 等ICLR 2023 · 被引用 94 次
相关 Paper
- ProReason: Multi-Modal Proactive Reasoning with Decoupled Eyesight and WisdomJingqi Zhou, Sheng Wang, Jingwei Dong, Kai Liu 等EMNLP 2025
- Towards a Mechanistic Interpretation of Multi-Step Reasoning Capabilities of Language ModelsYifan Hou, Jiaoda Li, Yu Fei, Alessandro Stolfo 等EMNLP 2023 · 被引用 2 次
- MetaMath: Bootstrap Your Own Mathematical Questions for Large Language ModelsLonghui Yu, Weisen Jiang, Han Shi, Jincheng Yu 等ICLR 2024 · 被引用 637 次
- R2-MultiOmnia: Leading Multilingual Multimodal Reasoning via Self-TrainingLeonardo Ranaldi, Federico Ranaldi, Giulia PucciACL 2025 · 被引用 9 次
- Selection-Inference: Exploiting Large Language Models for Interpretable Logical ReasoningAntonia Creswell, Murray Shanahan, Irina HigginsICLR 2023 · 被引用 110 次
