@inproceedings{lei-etal-2023-revisiting,
title = "Revisiting k-{NN} for Fine-tuning Pre-trained Language Models",
author = "Lei, Li and
Jing, Chen and
Botzhong, Tian and
Ningyu, Zhang",
editor = "Sun, Maosong and
Qin, Bing and
Qiu, Xipeng and
Jiang, Jing and
Han, Xianpei",
booktitle = "Proceedings of the 22nd Chinese National Conference on Computational Linguistics",
month = aug,
year = "2023",
address = "Harbin, China",
publisher = "Chinese Information Processing Society of China",
url = "https://aclanthology.org/2023.ccl-1.75",
pages = "889--897",
abstract = "{``}Pre-trained Language Models (PLMs), as parametric-based eager learners, have become thede-facto choice for current paradigms of Natural Language Processing (NLP). In contrast, k-Nearest-Neighbor (k-NN) classifiers, as the lazy learning paradigm, tend to mitigate over-fittingand isolated noise. In this paper, we revisit k-NN classifiers for augmenting the PLMs-based clas-sifiers. From the methodological level, we propose to adopt k-NN with textual representationsof PLMs in two steps: (1) Utilize k-NN as prior knowledge to calibrate the training process.(2) Linearly interpolate the probability distribution predicted by k-NN with that of the PLMs{'}classifier. At the heart of our approach is the implementation of k-NN-calibrated training, whichtreats predicted results as indicators for easy versus hard examples during the training process. From the perspective of the diversity of application scenarios, we conduct extensive experimentson fine-tuning, prompt-tuning paradigms and zero-shot, few-shot and fully-supervised settings,respectively, across eight diverse end-tasks. We hope our exploration will encourage the commu-nity to revisit the power of classical methods for efficient NLP1.{''}",
language = "English",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lei-etal-2023-revisiting">
<titleInfo>
<title>Revisiting k-NN for Fine-tuning Pre-trained Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Li</namePart>
<namePart type="family">Lei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chen</namePart>
<namePart type="family">Jing</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tian</namePart>
<namePart type="family">Botzhong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhang</namePart>
<namePart type="family">Ningyu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<language>
<languageTerm type="text">English</languageTerm>
<languageTerm type="code" authority="iso639-2b">eng</languageTerm>
</language>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 22nd Chinese National Conference on Computational Linguistics</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maosong</namePart>
<namePart type="family">Sun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bing</namePart>
<namePart type="family">Qin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xipeng</namePart>
<namePart type="family">Qiu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jing</namePart>
<namePart type="family">Jiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xianpei</namePart>
<namePart type="family">Han</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Chinese Information Processing Society of China</publisher>
<place>
<placeTerm type="text">Harbin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>“Pre-trained Language Models (PLMs), as parametric-based eager learners, have become thede-facto choice for current paradigms of Natural Language Processing (NLP). In contrast, k-Nearest-Neighbor (k-NN) classifiers, as the lazy learning paradigm, tend to mitigate over-fittingand isolated noise. In this paper, we revisit k-NN classifiers for augmenting the PLMs-based clas-sifiers. From the methodological level, we propose to adopt k-NN with textual representationsof PLMs in two steps: (1) Utilize k-NN as prior knowledge to calibrate the training process.(2) Linearly interpolate the probability distribution predicted by k-NN with that of the PLMs’classifier. At the heart of our approach is the implementation of k-NN-calibrated training, whichtreats predicted results as indicators for easy versus hard examples during the training process. From the perspective of the diversity of application scenarios, we conduct extensive experimentson fine-tuning, prompt-tuning paradigms and zero-shot, few-shot and fully-supervised settings,respectively, across eight diverse end-tasks. We hope our exploration will encourage the commu-nity to revisit the power of classical methods for efficient NLP1.”</abstract>
<identifier type="citekey">lei-etal-2023-revisiting</identifier>
<location>
<url>https://aclanthology.org/2023.ccl-1.75</url>
</location>
<part>
<date>2023-08</date>
<extent unit="page">
<start>889</start>
<end>897</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Revisiting k-NN for Fine-tuning Pre-trained Language Models
%A Lei, Li
%A Jing, Chen
%A Botzhong, Tian
%A Ningyu, Zhang
%Y Sun, Maosong
%Y Qin, Bing
%Y Qiu, Xipeng
%Y Jiang, Jing
%Y Han, Xianpei
%S Proceedings of the 22nd Chinese National Conference on Computational Linguistics
%D 2023
%8 August
%I Chinese Information Processing Society of China
%C Harbin, China
%G English
%F lei-etal-2023-revisiting
%X “Pre-trained Language Models (PLMs), as parametric-based eager learners, have become thede-facto choice for current paradigms of Natural Language Processing (NLP). In contrast, k-Nearest-Neighbor (k-NN) classifiers, as the lazy learning paradigm, tend to mitigate over-fittingand isolated noise. In this paper, we revisit k-NN classifiers for augmenting the PLMs-based clas-sifiers. From the methodological level, we propose to adopt k-NN with textual representationsof PLMs in two steps: (1) Utilize k-NN as prior knowledge to calibrate the training process.(2) Linearly interpolate the probability distribution predicted by k-NN with that of the PLMs’classifier. At the heart of our approach is the implementation of k-NN-calibrated training, whichtreats predicted results as indicators for easy versus hard examples during the training process. From the perspective of the diversity of application scenarios, we conduct extensive experimentson fine-tuning, prompt-tuning paradigms and zero-shot, few-shot and fully-supervised settings,respectively, across eight diverse end-tasks. We hope our exploration will encourage the commu-nity to revisit the power of classical methods for efficient NLP1.”
%U https://aclanthology.org/2023.ccl-1.75
%P 889-897
Markdown (Informal)
[Revisiting k-NN for Fine-tuning Pre-trained Language Models](https://aclanthology.org/2023.ccl-1.75) (Lei et al., CCL 2023)
ACL