<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-etal-2025-task">
    <titleInfo>
        <title>Task Calibration: Calibrating Large Language Models on Inference Tasks</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yingjie</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yun</namePart>
        <namePart type="family">Luo</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Xiaotian</namePart>
        <namePart type="family">Xie</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yue</namePart>
        <namePart type="family">Zhang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: ACL 2025</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Wanxiang</namePart>
            <namePart type="family">Che</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Joyce</namePart>
            <namePart type="family">Nabende</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ekaterina</namePart>
            <namePart type="family">Shutova</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Mohammad</namePart>
            <namePart type="given">Taher</namePart>
            <namePart type="family">Pilehvar</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Vienna, Austria</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-256-5</identifier>
    </relatedItem>
    <abstract>Large language models (LLMs) have exhibited impressive zero-shot performance on inference tasks. However, LLMs may suffer from spurious correlations between input texts and output labels, which limits LLMs’ ability to reason based purely on general language understanding. For example, in the natural language inference (NLI) task, LLMs may make predictions primarily based on premise or hypothesis, rather than both components. To address this problem that may lead to unexpected performance degradation, we propose task calibration (TC), a zero-shot and inference-only calibration method inspired by mutual information which recovers LLM performance through task reformulation. In NLI, TC encourages LLMs to reason based on both premise and hypothesis, while mitigating the models’ over-reliance on individual premise or hypothesis for inference. Experimental results show that TC achieves a substantial improvement on 13 different benchmarks in the zero-shot setup. We further validate the effectiveness of TC in few-shot setups and various natural language understanding tasks. Further analysis indicates that TC is also robust to prompt templates and has the potential to be integrated with other calibration methods. We publicly release our code to facilitate future research.</abstract>
    <identifier type="citekey">li-etal-2025-task</identifier>
    <identifier type="doi">10.18653/v1/2025.findings-acl.362</identifier>
    <location>
        <url>https://aclanthology.org/2025.findings-acl.362/</url>
    </location>
    <part>
        <date>2025-07</date>
        <extent unit="page">
            <start>6937</start>
            <end>6951</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Task Calibration: Calibrating Large Language Models on Inference Tasks
%A Li, Yingjie
%A Luo, Yun
%A Xie, Xiaotian
%A Zhang, Yue
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Findings of the Association for Computational Linguistics: ACL 2025
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-256-5
%F li-etal-2025-task
%X Large language models (LLMs) have exhibited impressive zero-shot performance on inference tasks. However, LLMs may suffer from spurious correlations between input texts and output labels, which limits LLMs’ ability to reason based purely on general language understanding. For example, in the natural language inference (NLI) task, LLMs may make predictions primarily based on premise or hypothesis, rather than both components. To address this problem that may lead to unexpected performance degradation, we propose task calibration (TC), a zero-shot and inference-only calibration method inspired by mutual information which recovers LLM performance through task reformulation. In NLI, TC encourages LLMs to reason based on both premise and hypothesis, while mitigating the models’ over-reliance on individual premise or hypothesis for inference. Experimental results show that TC achieves a substantial improvement on 13 different benchmarks in the zero-shot setup. We further validate the effectiveness of TC in few-shot setups and various natural language understanding tasks. Further analysis indicates that TC is also robust to prompt templates and has the potential to be integrated with other calibration methods. We publicly release our code to facilitate future research.
%R 10.18653/v1/2025.findings-acl.362
%U https://aclanthology.org/2025.findings-acl.362/
%U https://doi.org/10.18653/v1/2025.findings-acl.362
%P 6937-6951

Markdown (Informal)

[Task Calibration: Calibrating Large Language Models on Inference Tasks](https://aclanthology.org/2025.findings-acl.362/) (Li et al., Findings 2025)

• Task Calibration: Calibrating Large Language Models on Inference Tasks (Li et al., Findings 2025)

ACL

• Yingjie Li, Yun Luo, Xiaotian Xie, and Yue Zhang. 2025. Task Calibration: Calibrating Large Language Models on Inference Tasks. In Findings of the Association for Computational Linguistics: ACL 2025, pages 6937–6951, Vienna, Austria. Association for Computational Linguistics.