@inproceedings{xu-etal-2024-reading,
title = "Re-Reading Improves Reasoning in Large Language Models",
author = "Xu, Xiaohan and
Tao, Chongyang and
Shen, Tao and
Xu, Can and
Xu, Hongbo and
Long, Guodong and
Lou, Jian-Guang and
Ma, Shuai",
editor = "Al-Onaizan, Yaser and
Bansal, Mohit and
Chen, Yun-Nung",
booktitle = "Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.emnlp-main.871/",
doi = "10.18653/v1/2024.emnlp-main.871",
pages = "15549--15575",
abstract = "To enhance the reasoning capabilities of off-the-shelf Large Language Models (LLMs), we introduce a simple, yet general and effective prompting method, RE2, i.e., Re-Reading the question as input. Unlike most thought-eliciting prompting methods, such as Chain-of-Thought (CoT), which aim to elicit the reasoning process in the output, RE2 shifts the focus to the input by processing questions twice, thereby enhancing the understanding process. Consequently, RE2 demonstrates strong generality and compatibility with most thought-eliciting prompting methods, including CoT. Crucially, RE2 facilitates a {\textquotedblleft}bidirectional{\textquotedblright} encoding in unidirectional decoder-only LLMs because the first pass could provide global information for the second pass. We begin with a preliminary empirical study as the foundation of RE2, illustrating its potential to enable {\textquotedblleft}bidirectional{\textquotedblright} attention mechanisms. We then evaluate RE2 on extensive reasoning benchmarks across 14 datasets, spanning 112 experiments, to validate its effectiveness and generality. Our findings indicate that, with the exception of a few scenarios on vanilla ChatGPT, RE2 consistently enhances the reasoning performance of LLMs through a simple re-reading strategy. Further analyses reveal RE2`s adaptability, showing how it can be effectively integrated with different LLMs, thought-eliciting prompting, and ensemble strategies."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="xu-etal-2024-reading">
<titleInfo>
<title>Re-Reading Improves Reasoning in Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xiaohan</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chongyang</namePart>
<namePart type="family">Tao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tao</namePart>
<namePart type="family">Shen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Can</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hongbo</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Guodong</namePart>
<namePart type="family">Long</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jian-Guang</namePart>
<namePart type="family">Lou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shuai</namePart>
<namePart type="family">Ma</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>Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing</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>To enhance the reasoning capabilities of off-the-shelf Large Language Models (LLMs), we introduce a simple, yet general and effective prompting method, RE2, i.e., Re-Reading the question as input. Unlike most thought-eliciting prompting methods, such as Chain-of-Thought (CoT), which aim to elicit the reasoning process in the output, RE2 shifts the focus to the input by processing questions twice, thereby enhancing the understanding process. Consequently, RE2 demonstrates strong generality and compatibility with most thought-eliciting prompting methods, including CoT. Crucially, RE2 facilitates a “bidirectional” encoding in unidirectional decoder-only LLMs because the first pass could provide global information for the second pass. We begin with a preliminary empirical study as the foundation of RE2, illustrating its potential to enable “bidirectional” attention mechanisms. We then evaluate RE2 on extensive reasoning benchmarks across 14 datasets, spanning 112 experiments, to validate its effectiveness and generality. Our findings indicate that, with the exception of a few scenarios on vanilla ChatGPT, RE2 consistently enhances the reasoning performance of LLMs through a simple re-reading strategy. Further analyses reveal RE2‘s adaptability, showing how it can be effectively integrated with different LLMs, thought-eliciting prompting, and ensemble strategies.</abstract>
<identifier type="citekey">xu-etal-2024-reading</identifier>
<identifier type="doi">10.18653/v1/2024.emnlp-main.871</identifier>
<location>
<url>https://aclanthology.org/2024.emnlp-main.871/</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>15549</start>
<end>15575</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Re-Reading Improves Reasoning in Large Language Models
%A Xu, Xiaohan
%A Tao, Chongyang
%A Shen, Tao
%A Xu, Can
%A Xu, Hongbo
%A Long, Guodong
%A Lou, Jian-Guang
%A Ma, Shuai
%Y Al-Onaizan, Yaser
%Y Bansal, Mohit
%Y Chen, Yun-Nung
%S Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F xu-etal-2024-reading
%X To enhance the reasoning capabilities of off-the-shelf Large Language Models (LLMs), we introduce a simple, yet general and effective prompting method, RE2, i.e., Re-Reading the question as input. Unlike most thought-eliciting prompting methods, such as Chain-of-Thought (CoT), which aim to elicit the reasoning process in the output, RE2 shifts the focus to the input by processing questions twice, thereby enhancing the understanding process. Consequently, RE2 demonstrates strong generality and compatibility with most thought-eliciting prompting methods, including CoT. Crucially, RE2 facilitates a “bidirectional” encoding in unidirectional decoder-only LLMs because the first pass could provide global information for the second pass. We begin with a preliminary empirical study as the foundation of RE2, illustrating its potential to enable “bidirectional” attention mechanisms. We then evaluate RE2 on extensive reasoning benchmarks across 14 datasets, spanning 112 experiments, to validate its effectiveness and generality. Our findings indicate that, with the exception of a few scenarios on vanilla ChatGPT, RE2 consistently enhances the reasoning performance of LLMs through a simple re-reading strategy. Further analyses reveal RE2‘s adaptability, showing how it can be effectively integrated with different LLMs, thought-eliciting prompting, and ensemble strategies.
%R 10.18653/v1/2024.emnlp-main.871
%U https://aclanthology.org/2024.emnlp-main.871/
%U https://doi.org/10.18653/v1/2024.emnlp-main.871
%P 15549-15575
Markdown (Informal)
[Re-Reading Improves Reasoning in Large Language Models](https://aclanthology.org/2024.emnlp-main.871/) (Xu et al., EMNLP 2024)
ACL
- Xiaohan Xu, Chongyang Tao, Tao Shen, Can Xu, Hongbo Xu, Guodong Long, Jian-Guang Lou, and Shuai Ma. 2024. Re-Reading Improves Reasoning in Large Language Models. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 15549–15575, Miami, Florida, USA. Association for Computational Linguistics.