@inproceedings{chen-etal-2025-finereason,
title = "{F}ine{R}eason: Evaluating and Improving {LLM}s' Deliberate Reasoning through Reflective Puzzle Solving",
author = "Chen, Guizhen and
Xu, Weiwen and
Zhang, Hao and
Chan, Hou Pong and
Liu, Chaoqun and
Bing, Lidong and
Zhao, Deli and
Luu, Anh Tuan and
Rong, Yu",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.acl-long.333/",
doi = "10.18653/v1/2025.acl-long.333",
pages = "6685--6715",
ISBN = "979-8-89176-251-0",
abstract = "Many challenging reasoning tasks require not just rapid, intuitive responses, but a more deliberate, multi-step approach. Recent progress in large language models (LLMs) highlights an important shift from the ``System 1'' way of quick reactions to the ``System 2'' style of reflection-and-correction problem solving. However, current benchmarks heavily rely on the final-answer accuracy, leaving much of a model{'}s intermediate reasoning steps unexamined. This fails to assess the model{'}s ability to reflect and rectify mistakes within the reasoning process. To bridge this gap, we introduce FINEREASON, a logic-puzzle benchmark for systematic evaluation of LLMs' reasoning capabilities. Each puzzle can be decomposed into atomic steps, making it ideal for rigorous validation of intermediate correctness. Building on this, we introduce two tasks: state checking and state transition, for a comprehensive evaluation of how models assess the current situation and plan the next move. To support broader research, we also provide a puzzle training set aimed at enhancing general reasoning. We show that models trained on our state checking and transition data demonstrate gains in mathematical reasoning by up to 5.1{\%}."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="chen-etal-2025-finereason">
<titleInfo>
<title>FineReason: Evaluating and Improving LLMs’ Deliberate Reasoning through Reflective Puzzle Solving</title>
</titleInfo>
<name type="personal">
<namePart type="given">Guizhen</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Weiwen</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hao</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hou</namePart>
<namePart type="given">Pong</namePart>
<namePart type="family">Chan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chaoqun</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lidong</namePart>
<namePart type="family">Bing</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Deli</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anh</namePart>
<namePart type="given">Tuan</namePart>
<namePart type="family">Luu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yu</namePart>
<namePart type="family">Rong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-251-0</identifier>
</relatedItem>
<abstract>Many challenging reasoning tasks require not just rapid, intuitive responses, but a more deliberate, multi-step approach. Recent progress in large language models (LLMs) highlights an important shift from the “System 1” way of quick reactions to the “System 2” style of reflection-and-correction problem solving. However, current benchmarks heavily rely on the final-answer accuracy, leaving much of a model’s intermediate reasoning steps unexamined. This fails to assess the model’s ability to reflect and rectify mistakes within the reasoning process. To bridge this gap, we introduce FINEREASON, a logic-puzzle benchmark for systematic evaluation of LLMs’ reasoning capabilities. Each puzzle can be decomposed into atomic steps, making it ideal for rigorous validation of intermediate correctness. Building on this, we introduce two tasks: state checking and state transition, for a comprehensive evaluation of how models assess the current situation and plan the next move. To support broader research, we also provide a puzzle training set aimed at enhancing general reasoning. We show that models trained on our state checking and transition data demonstrate gains in mathematical reasoning by up to 5.1%.</abstract>
<identifier type="citekey">chen-etal-2025-finereason</identifier>
<identifier type="doi">10.18653/v1/2025.acl-long.333</identifier>
<location>
<url>https://aclanthology.org/2025.acl-long.333/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>6685</start>
<end>6715</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T FineReason: Evaluating and Improving LLMs’ Deliberate Reasoning through Reflective Puzzle Solving
%A Chen, Guizhen
%A Xu, Weiwen
%A Zhang, Hao
%A Chan, Hou Pong
%A Liu, Chaoqun
%A Bing, Lidong
%A Zhao, Deli
%A Luu, Anh Tuan
%A Rong, Yu
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-251-0
%F chen-etal-2025-finereason
%X Many challenging reasoning tasks require not just rapid, intuitive responses, but a more deliberate, multi-step approach. Recent progress in large language models (LLMs) highlights an important shift from the “System 1” way of quick reactions to the “System 2” style of reflection-and-correction problem solving. However, current benchmarks heavily rely on the final-answer accuracy, leaving much of a model’s intermediate reasoning steps unexamined. This fails to assess the model’s ability to reflect and rectify mistakes within the reasoning process. To bridge this gap, we introduce FINEREASON, a logic-puzzle benchmark for systematic evaluation of LLMs’ reasoning capabilities. Each puzzle can be decomposed into atomic steps, making it ideal for rigorous validation of intermediate correctness. Building on this, we introduce two tasks: state checking and state transition, for a comprehensive evaluation of how models assess the current situation and plan the next move. To support broader research, we also provide a puzzle training set aimed at enhancing general reasoning. We show that models trained on our state checking and transition data demonstrate gains in mathematical reasoning by up to 5.1%.
%R 10.18653/v1/2025.acl-long.333
%U https://aclanthology.org/2025.acl-long.333/
%U https://doi.org/10.18653/v1/2025.acl-long.333
%P 6685-6715
Markdown (Informal)
[FineReason: Evaluating and Improving LLMs’ Deliberate Reasoning through Reflective Puzzle Solving](https://aclanthology.org/2025.acl-long.333/) (Chen et al., ACL 2025)
ACL
- Guizhen Chen, Weiwen Xu, Hao Zhang, Hou Pong Chan, Chaoqun Liu, Lidong Bing, Deli Zhao, Anh Tuan Luu, and Yu Rong. 2025. FineReason: Evaluating and Improving LLMs’ Deliberate Reasoning through Reflective Puzzle Solving. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 6685–6715, Vienna, Austria. Association for Computational Linguistics.