@inproceedings{sicilia-alikhani-2025-adaptive,
title = "Adaptive {P}latt Scaling with Causal Interpretations for Self-Reflective Language Model Uncertainty Estimates",
author = "Sicilia, Anthony and
Alikhani, Malihe",
editor = "Christodoulopoulos, Christos and
Chakraborty, Tanmoy and
Rose, Carolyn and
Peng, Violet",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-emnlp.999/",
doi = "10.18653/v1/2025.findings-emnlp.999",
pages = "18414--18422",
ISBN = "979-8-89176-335-7",
abstract = "As large language models (LLMs) are consumed by more users and deployed in increasingly autonomous capacities, their ability to self-monitor and ask for human intervention is of vital importance. Underlying this capability are fundamental skills like self-reflection and expression of uncertainty. In this work, we provide a formal analysis of LLM self-reflection for uncertainty estimation, using domain adaptation theory to model the shift between base predictions and reflective judgments. We use this to motivate a temperature scaling algorithm that calibrates uncertainty using comparisons between base predictions and LLM self-reflections. We evaluate our approach on challenging question-answering tasks requiring reasoning, demonstrating that our methods can improve calibration of uncertainty estimates and also offer improvements in human interpretation. More broadly, this use case shows how domain adaptation presents a promising analytical tool for understanding the underlying statistical properties of LLM self-reflections."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="sicilia-alikhani-2025-adaptive">
<titleInfo>
<title>Adaptive Platt Scaling with Causal Interpretations for Self-Reflective Language Model Uncertainty Estimates</title>
</titleInfo>
<name type="personal">
<namePart type="given">Anthony</namePart>
<namePart type="family">Sicilia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Malihe</namePart>
<namePart type="family">Alikhani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Christos</namePart>
<namePart type="family">Christodoulopoulos</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tanmoy</namePart>
<namePart type="family">Chakraborty</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Carolyn</namePart>
<namePart type="family">Rose</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Violet</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-335-7</identifier>
</relatedItem>
<abstract>As large language models (LLMs) are consumed by more users and deployed in increasingly autonomous capacities, their ability to self-monitor and ask for human intervention is of vital importance. Underlying this capability are fundamental skills like self-reflection and expression of uncertainty. In this work, we provide a formal analysis of LLM self-reflection for uncertainty estimation, using domain adaptation theory to model the shift between base predictions and reflective judgments. We use this to motivate a temperature scaling algorithm that calibrates uncertainty using comparisons between base predictions and LLM self-reflections. We evaluate our approach on challenging question-answering tasks requiring reasoning, demonstrating that our methods can improve calibration of uncertainty estimates and also offer improvements in human interpretation. More broadly, this use case shows how domain adaptation presents a promising analytical tool for understanding the underlying statistical properties of LLM self-reflections.</abstract>
<identifier type="citekey">sicilia-alikhani-2025-adaptive</identifier>
<identifier type="doi">10.18653/v1/2025.findings-emnlp.999</identifier>
<location>
<url>https://aclanthology.org/2025.findings-emnlp.999/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>18414</start>
<end>18422</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Adaptive Platt Scaling with Causal Interpretations for Self-Reflective Language Model Uncertainty Estimates
%A Sicilia, Anthony
%A Alikhani, Malihe
%Y Christodoulopoulos, Christos
%Y Chakraborty, Tanmoy
%Y Rose, Carolyn
%Y Peng, Violet
%S Findings of the Association for Computational Linguistics: EMNLP 2025
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-335-7
%F sicilia-alikhani-2025-adaptive
%X As large language models (LLMs) are consumed by more users and deployed in increasingly autonomous capacities, their ability to self-monitor and ask for human intervention is of vital importance. Underlying this capability are fundamental skills like self-reflection and expression of uncertainty. In this work, we provide a formal analysis of LLM self-reflection for uncertainty estimation, using domain adaptation theory to model the shift between base predictions and reflective judgments. We use this to motivate a temperature scaling algorithm that calibrates uncertainty using comparisons between base predictions and LLM self-reflections. We evaluate our approach on challenging question-answering tasks requiring reasoning, demonstrating that our methods can improve calibration of uncertainty estimates and also offer improvements in human interpretation. More broadly, this use case shows how domain adaptation presents a promising analytical tool for understanding the underlying statistical properties of LLM self-reflections.
%R 10.18653/v1/2025.findings-emnlp.999
%U https://aclanthology.org/2025.findings-emnlp.999/
%U https://doi.org/10.18653/v1/2025.findings-emnlp.999
%P 18414-18422
Markdown (Informal)
[Adaptive Platt Scaling with Causal Interpretations for Self-Reflective Language Model Uncertainty Estimates](https://aclanthology.org/2025.findings-emnlp.999/) (Sicilia & Alikhani, Findings 2025)
ACL