@inproceedings{tripathy-2026-confidence,
title = "Confidence Drift During Task-Specific Language Model Fine-Tuning",
author = "Tripathy, Soham",
editor = {Aziz, Wilker and
Berant, Jonathan and
Eikema, Bryan and
Ilia, Evgenia and
de Marneffe, Marie-Catherine and
Plank, Barbara and
Shelmanov, Artem and
Swayamdipta, Swabha and
Tiedemann, J{\"o}rg and
Vazhentsev, Artem and
V{\'a}zquez, Ra{\'u}l and
Zerva, Chrysoula},
booktitle = "The Proceedings of the Third Workshop on Uncertainty-Aware {NLP}",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.uncertainlp-main.42/",
pages = "693--703",
ISBN = "979-8-89176-486-6",
abstract = "Task-specific fine-tuning can change not only what a language model predicts, but also how confident it is in those predictions. We study confidence and accuracy drifts throughout LoRA fine-tuning of seven decoder-only language models spanning multiple families and scales. Models are fine-tuned on AG News and evaluated at intermediate checkpoints on the source task and five untouched cross-task benchmarks. We find that cross-task accuracy generally remains stable or degrades, while confidence increases on both correct and incorrect predictions. Across many settings, confidence follows a non-monotonic trajectory: after an early transient and a relatively stable middle stage, it rises again late in training, coinciding with increasing train{--}validation NLL divergence. Our results show that accuracy-only evaluation can miss confidence drifts introduced by task-specific fine-tuning."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="tripathy-2026-confidence">
<titleInfo>
<title>Confidence Drift During Task-Specific Language Model Fine-Tuning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Soham</namePart>
<namePart type="family">Tripathy</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>The Proceedings of the Third Workshop on Uncertainty-Aware NLP</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wilker</namePart>
<namePart type="family">Aziz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jonathan</namePart>
<namePart type="family">Berant</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bryan</namePart>
<namePart type="family">Eikema</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Evgenia</namePart>
<namePart type="family">Ilia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marie-Catherine</namePart>
<namePart type="family">de Marneffe</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Barbara</namePart>
<namePart type="family">Plank</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Artem</namePart>
<namePart type="family">Shelmanov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Swabha</namePart>
<namePart type="family">Swayamdipta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jörg</namePart>
<namePart type="family">Tiedemann</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Artem</namePart>
<namePart type="family">Vazhentsev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Raúl</namePart>
<namePart type="family">Vázquez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chrysoula</namePart>
<namePart type="family">Zerva</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">979-8-89176-486-6</identifier>
</relatedItem>
<abstract>Task-specific fine-tuning can change not only what a language model predicts, but also how confident it is in those predictions. We study confidence and accuracy drifts throughout LoRA fine-tuning of seven decoder-only language models spanning multiple families and scales. Models are fine-tuned on AG News and evaluated at intermediate checkpoints on the source task and five untouched cross-task benchmarks. We find that cross-task accuracy generally remains stable or degrades, while confidence increases on both correct and incorrect predictions. Across many settings, confidence follows a non-monotonic trajectory: after an early transient and a relatively stable middle stage, it rises again late in training, coinciding with increasing train–validation NLL divergence. Our results show that accuracy-only evaluation can miss confidence drifts introduced by task-specific fine-tuning.</abstract>
<identifier type="citekey">tripathy-2026-confidence</identifier>
<location>
<url>https://aclanthology.org/2026.uncertainlp-main.42/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>693</start>
<end>703</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Confidence Drift During Task-Specific Language Model Fine-Tuning
%A Tripathy, Soham
%Y Aziz, Wilker
%Y Berant, Jonathan
%Y Eikema, Bryan
%Y Ilia, Evgenia
%Y de Marneffe, Marie-Catherine
%Y Plank, Barbara
%Y Shelmanov, Artem
%Y Swayamdipta, Swabha
%Y Tiedemann, Jörg
%Y Vazhentsev, Artem
%Y Vázquez, Raúl
%Y Zerva, Chrysoula
%S The Proceedings of the Third Workshop on Uncertainty-Aware NLP
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-486-6
%F tripathy-2026-confidence
%X Task-specific fine-tuning can change not only what a language model predicts, but also how confident it is in those predictions. We study confidence and accuracy drifts throughout LoRA fine-tuning of seven decoder-only language models spanning multiple families and scales. Models are fine-tuned on AG News and evaluated at intermediate checkpoints on the source task and five untouched cross-task benchmarks. We find that cross-task accuracy generally remains stable or degrades, while confidence increases on both correct and incorrect predictions. Across many settings, confidence follows a non-monotonic trajectory: after an early transient and a relatively stable middle stage, it rises again late in training, coinciding with increasing train–validation NLL divergence. Our results show that accuracy-only evaluation can miss confidence drifts introduced by task-specific fine-tuning.
%U https://aclanthology.org/2026.uncertainlp-main.42/
%P 693-703
Markdown (Informal)
[Confidence Drift During Task-Specific Language Model Fine-Tuning](https://aclanthology.org/2026.uncertainlp-main.42/) (Tripathy, UncertaiNLP 2026)
ACL