@inproceedings{kumari-etal-2024-end,
title = "An End-to-End Submodular Framework for Data-Efficient In-Context Learning",
author = "Kumari, Lilly and
Wang, Shengjie and
Das, Arnav and
Zhou, Tianyi and
Bilmes, Jeff",
editor = "Duh, Kevin and
Gomez, Helena and
Bethard, Steven",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2024",
month = jun,
year = "2024",
address = "Mexico City, Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-naacl.209",
doi = "10.18653/v1/2024.findings-naacl.209",
pages = "3293--3308",
abstract = "Recent advancements in natural language tasks leverage the emergent In-Context Learning (ICL) ability of pretrained Large Language Models (LLMs). ICL enables LLMs to perform new tasks by utilizing a limited number of input-output examples as prompts. While ICL circumvents the costly step of finetuning LLMs, its effectiveness is heavily dependent on the quality and ordering of provided examples (called exemplars). In this work, we propose a two-stage data-efficient framework $\textit{Div-S3}$ for exemplar selection for ICL. The first stage focuses on data annotation and employs a pool-based active learning approach to select a set of $\textit{Div}$erse and informative exemplars from the target tasks{'} unlabeled pool. Given a test input/query, the second stage uses Submodular Span Summarization ($\textit{S3}$) to select the most relevant and non-redundant exemplars from the annotated pool of a limited budget. On 7 different NLP datasets and 5 LLMs of varying complexities, we show $\textit{Div-S3}$ outperforms (1) existing active learning-based methods for data annotation for ICL and (2) similarity-based methods for test query-specific exemplars retrieval.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kumari-etal-2024-end">
<titleInfo>
<title>An End-to-End Submodular Framework for Data-Efficient In-Context Learning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Lilly</namePart>
<namePart type="family">Kumari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shengjie</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Arnav</namePart>
<namePart type="family">Das</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianyi</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jeff</namePart>
<namePart type="family">Bilmes</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-06</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: NAACL 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Kevin</namePart>
<namePart type="family">Duh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Helena</namePart>
<namePart type="family">Gomez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Steven</namePart>
<namePart type="family">Bethard</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Mexico City, Mexico</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Recent advancements in natural language tasks leverage the emergent In-Context Learning (ICL) ability of pretrained Large Language Models (LLMs). ICL enables LLMs to perform new tasks by utilizing a limited number of input-output examples as prompts. While ICL circumvents the costly step of finetuning LLMs, its effectiveness is heavily dependent on the quality and ordering of provided examples (called exemplars). In this work, we propose a two-stage data-efficient framework Div-S3 for exemplar selection for ICL. The first stage focuses on data annotation and employs a pool-based active learning approach to select a set of Diverse and informative exemplars from the target tasks’ unlabeled pool. Given a test input/query, the second stage uses Submodular Span Summarization (S3) to select the most relevant and non-redundant exemplars from the annotated pool of a limited budget. On 7 different NLP datasets and 5 LLMs of varying complexities, we show Div-S3 outperforms (1) existing active learning-based methods for data annotation for ICL and (2) similarity-based methods for test query-specific exemplars retrieval.</abstract>
<identifier type="citekey">kumari-etal-2024-end</identifier>
<identifier type="doi">10.18653/v1/2024.findings-naacl.209</identifier>
<location>
<url>https://aclanthology.org/2024.findings-naacl.209</url>
</location>
<part>
<date>2024-06</date>
<extent unit="page">
<start>3293</start>
<end>3308</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T An End-to-End Submodular Framework for Data-Efficient In-Context Learning
%A Kumari, Lilly
%A Wang, Shengjie
%A Das, Arnav
%A Zhou, Tianyi
%A Bilmes, Jeff
%Y Duh, Kevin
%Y Gomez, Helena
%Y Bethard, Steven
%S Findings of the Association for Computational Linguistics: NAACL 2024
%D 2024
%8 June
%I Association for Computational Linguistics
%C Mexico City, Mexico
%F kumari-etal-2024-end
%X Recent advancements in natural language tasks leverage the emergent In-Context Learning (ICL) ability of pretrained Large Language Models (LLMs). ICL enables LLMs to perform new tasks by utilizing a limited number of input-output examples as prompts. While ICL circumvents the costly step of finetuning LLMs, its effectiveness is heavily dependent on the quality and ordering of provided examples (called exemplars). In this work, we propose a two-stage data-efficient framework Div-S3 for exemplar selection for ICL. The first stage focuses on data annotation and employs a pool-based active learning approach to select a set of Diverse and informative exemplars from the target tasks’ unlabeled pool. Given a test input/query, the second stage uses Submodular Span Summarization (S3) to select the most relevant and non-redundant exemplars from the annotated pool of a limited budget. On 7 different NLP datasets and 5 LLMs of varying complexities, we show Div-S3 outperforms (1) existing active learning-based methods for data annotation for ICL and (2) similarity-based methods for test query-specific exemplars retrieval.
%R 10.18653/v1/2024.findings-naacl.209
%U https://aclanthology.org/2024.findings-naacl.209
%U https://doi.org/10.18653/v1/2024.findings-naacl.209
%P 3293-3308
Markdown (Informal)
[An End-to-End Submodular Framework for Data-Efficient In-Context Learning](https://aclanthology.org/2024.findings-naacl.209) (Kumari et al., Findings 2024)
ACL