@inproceedings{sun-etal-2024-unleashing,
title = "Unleashing the Potential of Large Language Models through Spectral Modulation",
author = "Sun, Peng and
Zhu, Yao and
Zhang, Yunjian and
Yan, Xiu and
Wang, Zizhe and
Ji, Xiangyang",
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.224/",
doi = "10.18653/v1/2024.findings-emnlp.224",
pages = "3892--3911",
abstract = "Large Language Models (LLMs) have demonstrated impressive capabilities across various domains, garnering significant attention from both academia and industry. However, enhancing the performance of LLMs typically requires scaling up model sizes or fine-tuning with additional datasets, which results in substantial computational costs. This paper poses an intriguing question: Can we improve the performance of LLMs without additional training? Drawing inspiration from signal processing principles, which suggest that noise often resides in high-frequency components while low-frequency components carry the essence of signals, we propose uncovering untapped potential in LLMs from a frequency perspective. We hypothesize that the high-frequency components in the weight matrices of LLMs' linear layers may conceal noise that interferes with predictive accuracy. Therefore, we propose conducting spectral modulation in the parameter space of LLMs, which can seamlessly integrate with various models in a plug-and-play manner. Extensive experiments have demonstrated the superiority of our approach, with spectral modulation yielding an average performance improvement of up to 10.12{\%}."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="sun-etal-2024-unleashing">
<titleInfo>
<title>Unleashing the Potential of Large Language Models through Spectral Modulation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Peng</namePart>
<namePart type="family">Sun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yao</namePart>
<namePart type="family">Zhu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yunjian</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiu</namePart>
<namePart type="family">Yan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zizhe</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiangyang</namePart>
<namePart type="family">Ji</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>Large Language Models (LLMs) have demonstrated impressive capabilities across various domains, garnering significant attention from both academia and industry. However, enhancing the performance of LLMs typically requires scaling up model sizes or fine-tuning with additional datasets, which results in substantial computational costs. This paper poses an intriguing question: Can we improve the performance of LLMs without additional training? Drawing inspiration from signal processing principles, which suggest that noise often resides in high-frequency components while low-frequency components carry the essence of signals, we propose uncovering untapped potential in LLMs from a frequency perspective. We hypothesize that the high-frequency components in the weight matrices of LLMs’ linear layers may conceal noise that interferes with predictive accuracy. Therefore, we propose conducting spectral modulation in the parameter space of LLMs, which can seamlessly integrate with various models in a plug-and-play manner. Extensive experiments have demonstrated the superiority of our approach, with spectral modulation yielding an average performance improvement of up to 10.12%.</abstract>
<identifier type="citekey">sun-etal-2024-unleashing</identifier>
<identifier type="doi">10.18653/v1/2024.findings-emnlp.224</identifier>
<location>
<url>https://aclanthology.org/2024.findings-emnlp.224/</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>3892</start>
<end>3911</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Unleashing the Potential of Large Language Models through Spectral Modulation
%A Sun, Peng
%A Zhu, Yao
%A Zhang, Yunjian
%A Yan, Xiu
%A Wang, Zizhe
%A Ji, Xiangyang
%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 sun-etal-2024-unleashing
%X Large Language Models (LLMs) have demonstrated impressive capabilities across various domains, garnering significant attention from both academia and industry. However, enhancing the performance of LLMs typically requires scaling up model sizes or fine-tuning with additional datasets, which results in substantial computational costs. This paper poses an intriguing question: Can we improve the performance of LLMs without additional training? Drawing inspiration from signal processing principles, which suggest that noise often resides in high-frequency components while low-frequency components carry the essence of signals, we propose uncovering untapped potential in LLMs from a frequency perspective. We hypothesize that the high-frequency components in the weight matrices of LLMs’ linear layers may conceal noise that interferes with predictive accuracy. Therefore, we propose conducting spectral modulation in the parameter space of LLMs, which can seamlessly integrate with various models in a plug-and-play manner. Extensive experiments have demonstrated the superiority of our approach, with spectral modulation yielding an average performance improvement of up to 10.12%.
%R 10.18653/v1/2024.findings-emnlp.224
%U https://aclanthology.org/2024.findings-emnlp.224/
%U https://doi.org/10.18653/v1/2024.findings-emnlp.224
%P 3892-3911
Markdown (Informal)
[Unleashing the Potential of Large Language Models through Spectral Modulation](https://aclanthology.org/2024.findings-emnlp.224/) (Sun et al., Findings 2024)
ACL