@inproceedings{singh-etal-2026-kite,
title = "{KITE}: Kernelized and Information Theoretic Exemplars for In-Context Learning",
author = "Singh, Vaibhav and
Ghosal, Soumya Suvra and
Joshua, Kapu Nirmal and
Pal, Soumyabrata and
Ray Chowdhury, Sayak",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1667/",
pages = "35547--35568",
ISBN = "979-8-89176-462-0",
abstract = "In-context learning (ICL) has emerged as a powerful paradigm for adapting large language models (LLMs) to new, data-scarce tasks using only a few carefully selected task-specific examples embedded in the prompt. However, given the limited context size of LLMs, a fundamental question arises: Which examples should be selected to maximize performance on a given user query? While nearest-neighbor-based methods have been widely adopted for this purpose, they suffer from well-known drawbacks in high-dimensional embedding spaces, including poor generalization and limited diversity. In this work, we study example selection in ICL from a principled, information-theoretic perspective. We frame the example selection task as a query-specific high-dimensional optimization problem: selecting a subset of exemplars from a larger example bank to minimize prediction error for a given query. To solve this, we first derive a principled surrogate objective that is approximately submodular, enabling the use of a greedy algorithm with an approximation guarantee. We further enhance our method by introducing an optimal design-based regularizer to encourage diversity in the selected examples. Empirically, we demonstrate significant improvements over standard retrieval methods across a suite of classification tasks, highlighting the benefits of structure-aware, diverse example selection for ICL in real-world, label-scarce scenarios."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="singh-etal-2026-kite">
<titleInfo>
<title>KITE: Kernelized and Information Theoretic Exemplars for In-Context Learning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Vaibhav</namePart>
<namePart type="family">Singh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Soumya</namePart>
<namePart type="given">Suvra</namePart>
<namePart type="family">Ghosal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kapu</namePart>
<namePart type="given">Nirmal</namePart>
<namePart type="family">Joshua</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Soumyabrata</namePart>
<namePart type="family">Pal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sayak</namePart>
<namePart type="family">Ray Chowdhury</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>In-context learning (ICL) has emerged as a powerful paradigm for adapting large language models (LLMs) to new, data-scarce tasks using only a few carefully selected task-specific examples embedded in the prompt. However, given the limited context size of LLMs, a fundamental question arises: Which examples should be selected to maximize performance on a given user query? While nearest-neighbor-based methods have been widely adopted for this purpose, they suffer from well-known drawbacks in high-dimensional embedding spaces, including poor generalization and limited diversity. In this work, we study example selection in ICL from a principled, information-theoretic perspective. We frame the example selection task as a query-specific high-dimensional optimization problem: selecting a subset of exemplars from a larger example bank to minimize prediction error for a given query. To solve this, we first derive a principled surrogate objective that is approximately submodular, enabling the use of a greedy algorithm with an approximation guarantee. We further enhance our method by introducing an optimal design-based regularizer to encourage diversity in the selected examples. Empirically, we demonstrate significant improvements over standard retrieval methods across a suite of classification tasks, highlighting the benefits of structure-aware, diverse example selection for ICL in real-world, label-scarce scenarios.</abstract>
<identifier type="citekey">singh-etal-2026-kite</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1667/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>35547</start>
<end>35568</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T KITE: Kernelized and Information Theoretic Exemplars for In-Context Learning
%A Singh, Vaibhav
%A Ghosal, Soumya Suvra
%A Joshua, Kapu Nirmal
%A Pal, Soumyabrata
%A Ray Chowdhury, Sayak
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F singh-etal-2026-kite
%X In-context learning (ICL) has emerged as a powerful paradigm for adapting large language models (LLMs) to new, data-scarce tasks using only a few carefully selected task-specific examples embedded in the prompt. However, given the limited context size of LLMs, a fundamental question arises: Which examples should be selected to maximize performance on a given user query? While nearest-neighbor-based methods have been widely adopted for this purpose, they suffer from well-known drawbacks in high-dimensional embedding spaces, including poor generalization and limited diversity. In this work, we study example selection in ICL from a principled, information-theoretic perspective. We frame the example selection task as a query-specific high-dimensional optimization problem: selecting a subset of exemplars from a larger example bank to minimize prediction error for a given query. To solve this, we first derive a principled surrogate objective that is approximately submodular, enabling the use of a greedy algorithm with an approximation guarantee. We further enhance our method by introducing an optimal design-based regularizer to encourage diversity in the selected examples. Empirically, we demonstrate significant improvements over standard retrieval methods across a suite of classification tasks, highlighting the benefits of structure-aware, diverse example selection for ICL in real-world, label-scarce scenarios.
%U https://aclanthology.org/2026.findings-emnlp.1667/
%P 35547-35568
Markdown (Informal)
[KITE: Kernelized and Information Theoretic Exemplars for In-Context Learning](https://aclanthology.org/2026.findings-emnlp.1667/) (Singh et al., Findings 2026)
ACL