Towards Reasoning in Large Language Models: A Survey
Jie Huang Kevin Chen-Chuan Chang Department of Computer Science, University of Illinois at Urbana-Champaign {jeffhj,
Abstract
Reasoning is a fundamental aspect of human intelligence that plays a crucial role in activities such as problem solving, decision making, and critical thinking. In recent years, large language models (LLMs) have made significant progress in natural language processing, and there is observation that these models may exhibit reasoning abilities when they are sufficiently large. However, it is not yet clear to what extent LLMs are capable of reasoning. This paper provides a comprehensive overview of the current state of knowledge on reasoning in LLMs, including techniques for improving and eliciting reasoning in these models, methods and benchmarks for evaluating reasoning abilities, findings and implications of previous research in this field, and suggestions on future directions. Our aim is to provide a detailed and up-to-date review of this topic and stimulate meaningful discussion and future work.111Paperlist can be found at https://github.com/jeffhj/LM-reasoning.
中文速览
大型语言模型（LLM）在数学推理、常识问答等任务上表现亮眼，但它们究竟具备多强的推理能力、是否真正在"推理"，学界至今众说纷纭。这篇综述系统梳理了当前提升和激发LLM推理能力的主要技术——从思维链提示（chain-of-thought prompting）、自洽性解码到各类微调方法——并整理了用于评估推理的主流数据集与评测框架。研究发现，通过让模型在给出答案前先生成中间推理步骤，LLM在算术、符号推理和常识推理等任务上的表现可以显著提升，但在需要严格规划或复杂多步推理的场景下仍存在明显短板。这项工作为研究者提供了一份全面的领域地图，有助于厘清LLM推理能力的边界，并为未来如何让AI真正具备可靠推理能力指明了方向。
原文 arXiv:2212.10403;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2212.10403v2