Towards Self-Improvement of LLMs via MCTS: Leveraging Stepwise Knowledge with Curriculum Preference Learning

• URL: http://arxiv.org/abs/2410.06508v1
• Date: Wed, 9 Oct 2024 03:20:02 GMT
• Title: Towards Self-Improvement of LLMs via MCTS: Leveraging Stepwise Knowledge with Curriculum Preference Learning
• Authors: Xiyao Wang, Linfeng Song, Ye Tian, Dian Yu, Baolin Peng, Haitao Mi, Furong Huang, Dong Yu,
• Abstract summary: Monte Carlo Tree Search (MCTS) has emerged as a powerful technique for enhancing the reasoning capabilities of LLMs. Existing distillation methods underutilize the rich trajectory information generated by MCTS. We propose AlphaLLM-CPL, a novel pairwise training framework that enables LLMs to self-improve through MCTS behavior distillation.
• Score: 70.16816087320585
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Monte Carlo Tree Search (MCTS) has recently emerged as a powerful technique for enhancing the reasoning capabilities of LLMs. Techniques such as SFT or DPO have enabled LLMs to distill high-quality behaviors from MCTS, improving their reasoning performance. However, existing distillation methods underutilize the rich trajectory information generated by MCTS, limiting the potential for improvements in LLM reasoning. In this paper, we propose AlphaLLM-CPL, a novel pairwise training framework that enables LLMs to self-improve through MCTS behavior distillation. AlphaLLM-CPL efficiently leverages MCTS trajectories via two key innovations: (1) AlphaLLM-CPL constructs stepwise trajectory pairs from child nodes sharing the same parent in the search tree, providing step-level information for more effective MCTS behavior distillation. (2) AlphaLLM-CPL introduces curriculum preference learning, dynamically adjusting the training sequence of trajectory pairs in each offline training epoch to prioritize critical learning steps and mitigate overfitting. Experimental results on mathematical reasoning tasks demonstrate that AlphaLLM-CPL significantly outperforms previous MCTS behavior distillation methods, substantially boosting the reasoning capabilities of LLMs.

Related papers

• Teaching LLMs According to Their Aptitude: Adaptive Reasoning for Mathematical Problem Solving [55.895917967408586]
  Existing approaches to mathematical reasoning with large language models rely on Chain-of-Thought (CoT) for generalizability or Tool-Integrated Reasoning (TIR) for precise computation. We propose TATA (Teaching LLMs According to Their Aptitude), an adaptive framework that enables LLMs to personalize their reasoning strategy spontaneously.
  arXiv  Detail & Related papers  (2025-02-17T16:56:23Z)
• Satori: Reinforcement Learning with Chain-of-Action-Thought Enhances LLM Reasoning via Autoregressive Search [57.28671084993782]
  Large language models (LLMs) have demonstrated remarkable reasoning capabilities across diverse domains. Recent studies have shown that increasing test-time computation enhances LLMs' reasoning capabilities. We propose a two-stage training paradigm: 1) a small-scale format tuning stage to internalize the COAT reasoning format and 2) a large-scale self-improvement stage leveraging reinforcement learning.
  arXiv  Detail & Related papers  (2025-02-04T17:26:58Z)
• Enhancing Reasoning through Process Supervision with Monte Carlo Tree Search [2.1637240640145343]
  Large language models (LLMs) have demonstrated their remarkable capacity across a variety of tasks. To improve LLMs' reasoning ability, process supervision has proven to be better than outcome supervision. In this work, we study using Monte Carlo Tree Search (MCTS) to generate process supervision data with LLMs themselves for training them.
  arXiv  Detail & Related papers  (2025-01-02T12:09:17Z)
• CoMMIT: Coordinated Instruction Tuning for Multimodal Large Language Models [68.64605538559312]
  In this paper, we analyze the MLLM instruction tuning from both theoretical and empirical perspectives. Inspired by our findings, we propose a measurement to quantitatively evaluate the learning balance. In addition, we introduce an auxiliary loss regularization method to promote updating of the generation distribution of MLLMs.
  arXiv  Detail & Related papers  (2024-07-29T23:18:55Z)
• Q*: Improving Multi-step Reasoning for LLMs with Deliberative Planning [53.6472920229013]
  Large Language Models (LLMs) have demonstrated impressive capability in many natural language tasks. LLMs are prone to produce errors, hallucinations and inconsistent statements when performing multi-step reasoning. We introduce Q*, a framework for guiding LLMs decoding process with deliberative planning.
  arXiv  Detail & Related papers  (2024-06-20T13:08:09Z)
• Distilling Instruction-following Abilities of Large Language Models with Task-aware Curriculum Planning [12.651588927599441]
  Instruction tuning aims to align large language models with open-domain instructions and human-preferred responses. We introduce Task-Aware Curriculum Planning for Instruction Refinement (TAPIR) to select instructions that are difficult for a student LLM to follow. To balance the student's capabilities, task distributions in training sets are adjusted with responses automatically refined according to their corresponding tasks.
  arXiv  Detail & Related papers  (2024-05-22T08:38:26Z)
• Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizing [56.75702900542643]
  We introduce AlphaLLM for the self-improvements of Large Language Models. It integrates Monte Carlo Tree Search (MCTS) with LLMs to establish a self-improving loop. Our experimental results show that AlphaLLM significantly enhances the performance of LLMs without additional annotations.
  arXiv  Detail & Related papers  (2024-04-18T15:21:34Z)
• Knowledgeable Agents by Offline Reinforcement Learning from Large Language Model Rollouts [10.929547354171723]
  This paper introduces Knowledgeable Agents from Language Model Rollouts (KALM) It extracts knowledge from large language models (LLMs) in the form of imaginary rollouts that can be easily learned by the agent through offline reinforcement learning methods. It achieves a success rate of 46% in executing tasks with unseen goals, substantially surpassing the 26% success rate achieved by baseline methods.
  arXiv  Detail & Related papers  (2024-04-14T13:19:40Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.