@inproceedings{chiang-yogatama-2026-pelican,
title = "Pelican Soup Framework: A Theoretical Framework for Language Model Capabilities",
author = "Chiang, Ting-Rui and
Yogatama, Dani",
editor = "Demberg, Vera and
Inui, Kentaro and
Marquez, Llu{\'i}s",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EACL} 2026",
month = mar,
year = "2026",
address = "Rabat, Morocco",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-eacl.23/",
pages = "443--464",
ISBN = "979-8-89176-386-9",
abstract = "In this work, we propose a simple theoretical framework, Pelican Soup, aiming to better understand how pretraining allows LLMs to (1) generalize to unseen instructions and (2) perform in-context learning, even when the verbalizers are irrelevant to the task. To this end, in our framework, we introduce the notion of ``knowledge base'' and ``reference-sense association'' and a simple formalism for natural language processing tasks. Our framework demonstrates how linguistic, psychology, and philosophy studies can inform our understanding of the language model and is connected to several other existing theoretical results. As an illustration of the usage of our framework, we derive a bound on in-context learning loss with our framework. Finally, we support our framework with empirical experiments and provide possible future research directions."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="chiang-yogatama-2026-pelican">
<titleInfo>
<title>Pelican Soup Framework: A Theoretical Framework for Language Model Capabilities</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ting-Rui</namePart>
<namePart type="family">Chiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dani</namePart>
<namePart type="family">Yogatama</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-03</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Vera</namePart>
<namePart type="family">Demberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kentaro</namePart>
<namePart type="family">Inui</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lluís</namePart>
<namePart type="family">Marquez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Rabat, Morocco</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-386-9</identifier>
</relatedItem>
<abstract>In this work, we propose a simple theoretical framework, Pelican Soup, aiming to better understand how pretraining allows LLMs to (1) generalize to unseen instructions and (2) perform in-context learning, even when the verbalizers are irrelevant to the task. To this end, in our framework, we introduce the notion of “knowledge base” and “reference-sense association” and a simple formalism for natural language processing tasks. Our framework demonstrates how linguistic, psychology, and philosophy studies can inform our understanding of the language model and is connected to several other existing theoretical results. As an illustration of the usage of our framework, we derive a bound on in-context learning loss with our framework. Finally, we support our framework with empirical experiments and provide possible future research directions.</abstract>
<identifier type="citekey">chiang-yogatama-2026-pelican</identifier>
<location>
<url>https://aclanthology.org/2026.findings-eacl.23/</url>
</location>
<part>
<date>2026-03</date>
<extent unit="page">
<start>443</start>
<end>464</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Pelican Soup Framework: A Theoretical Framework for Language Model Capabilities
%A Chiang, Ting-Rui
%A Yogatama, Dani
%Y Demberg, Vera
%Y Inui, Kentaro
%Y Marquez, Lluís
%S Findings of the Association for Computational Linguistics: EACL 2026
%D 2026
%8 March
%I Association for Computational Linguistics
%C Rabat, Morocco
%@ 979-8-89176-386-9
%F chiang-yogatama-2026-pelican
%X In this work, we propose a simple theoretical framework, Pelican Soup, aiming to better understand how pretraining allows LLMs to (1) generalize to unseen instructions and (2) perform in-context learning, even when the verbalizers are irrelevant to the task. To this end, in our framework, we introduce the notion of “knowledge base” and “reference-sense association” and a simple formalism for natural language processing tasks. Our framework demonstrates how linguistic, psychology, and philosophy studies can inform our understanding of the language model and is connected to several other existing theoretical results. As an illustration of the usage of our framework, we derive a bound on in-context learning loss with our framework. Finally, we support our framework with empirical experiments and provide possible future research directions.
%U https://aclanthology.org/2026.findings-eacl.23/
%P 443-464
Markdown (Informal)
[Pelican Soup Framework: A Theoretical Framework for Language Model Capabilities](https://aclanthology.org/2026.findings-eacl.23/) (Chiang & Yogatama, Findings 2026)
ACL