@inproceedings{hwang-etal-2024-self,
title = "Self-Explore: Enhancing Mathematical Reasoning in Language Models with Fine-grained Rewards",
author = "Hwang, Hyeonbin and
Kim, Doyoung and
Kim, Seungone and
Ye, Seonghyeon and
Seo, Minjoon",
editor = "Al-Onaizan, Yaser and
Bansal, Mohit and
Chen, Yun-Nung",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-emnlp.78/",
doi = "10.18653/v1/2024.findings-emnlp.78",
pages = "1444--1466",
abstract = "Training on large amounts of rationales (i.e., CoT Fine-tuning) has been found effective for improving mathematical reasoning of large language models (LLMs). However, acquiring human-authored solutions or augmenting rationales from proprietary models is costly and not scalable. In this paper, we study the problem of whether LLMs could self-improve mathematical reasoning capabilities. To this end, we propose Self-Explore, where the LLM is tasked to explore the first wrong step (i.e., the first pit) within the rationale and use such signals as fine-grained rewards for further improvement. On the GSM8K and MATH test set, Self-Explore achieves 11.57{\%} and 2.89{\%} improvement on average across three LLMs compared to supervised fine-tuning (SFT). Our code is available here]9."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hwang-etal-2024-self">
<titleInfo>
<title>Self-Explore: Enhancing Mathematical Reasoning in Language Models with Fine-grained Rewards</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hyeonbin</namePart>
<namePart type="family">Hwang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Doyoung</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Seungone</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Seonghyeon</namePart>
<namePart type="family">Ye</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Minjoon</namePart>
<namePart type="family">Seo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yaser</namePart>
<namePart type="family">Al-Onaizan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohit</namePart>
<namePart type="family">Bansal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yun-Nung</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Miami, Florida, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Training on large amounts of rationales (i.e., CoT Fine-tuning) has been found effective for improving mathematical reasoning of large language models (LLMs). However, acquiring human-authored solutions or augmenting rationales from proprietary models is costly and not scalable. In this paper, we study the problem of whether LLMs could self-improve mathematical reasoning capabilities. To this end, we propose Self-Explore, where the LLM is tasked to explore the first wrong step (i.e., the first pit) within the rationale and use such signals as fine-grained rewards for further improvement. On the GSM8K and MATH test set, Self-Explore achieves 11.57% and 2.89% improvement on average across three LLMs compared to supervised fine-tuning (SFT). Our code is available here]9.</abstract>
<identifier type="citekey">hwang-etal-2024-self</identifier>
<identifier type="doi">10.18653/v1/2024.findings-emnlp.78</identifier>
<location>
<url>https://aclanthology.org/2024.findings-emnlp.78/</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>1444</start>
<end>1466</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Self-Explore: Enhancing Mathematical Reasoning in Language Models with Fine-grained Rewards
%A Hwang, Hyeonbin
%A Kim, Doyoung
%A Kim, Seungone
%A Ye, Seonghyeon
%A Seo, Minjoon
%Y Al-Onaizan, Yaser
%Y Bansal, Mohit
%Y Chen, Yun-Nung
%S Findings of the Association for Computational Linguistics: EMNLP 2024
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F hwang-etal-2024-self
%X Training on large amounts of rationales (i.e., CoT Fine-tuning) has been found effective for improving mathematical reasoning of large language models (LLMs). However, acquiring human-authored solutions or augmenting rationales from proprietary models is costly and not scalable. In this paper, we study the problem of whether LLMs could self-improve mathematical reasoning capabilities. To this end, we propose Self-Explore, where the LLM is tasked to explore the first wrong step (i.e., the first pit) within the rationale and use such signals as fine-grained rewards for further improvement. On the GSM8K and MATH test set, Self-Explore achieves 11.57% and 2.89% improvement on average across three LLMs compared to supervised fine-tuning (SFT). Our code is available here]9.
%R 10.18653/v1/2024.findings-emnlp.78
%U https://aclanthology.org/2024.findings-emnlp.78/
%U https://doi.org/10.18653/v1/2024.findings-emnlp.78
%P 1444-1466
Markdown (Informal)
[Self-Explore: Enhancing Mathematical Reasoning in Language Models with Fine-grained Rewards](https://aclanthology.org/2024.findings-emnlp.78/) (Hwang et al., Findings 2024)
ACL