@inproceedings{liu-etal-2026-semi,
title = "Semi-Supervised Learning to Improve Detection of Factual and Causal Errors in Language Models' Explanations",
author = "Liu, Mingxuan and
Yuan, Han and
Wu, Yilin and
Zhang, Li and
Ma, Zheng",
editor = "Wang, Yimu and
Choi, Yee Man and
Wu, Di and
Ouyang, Siqi and
Azadani, Mozhgan Nasr and
Fung, Yi R. (May)",
booktitle = "Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently ({G}round{LM} 2026)",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.groundlm-1.11/",
pages = "141--147",
ISBN = "979-8-89176-478-1",
abstract = "Language models (LMs) are increasingly used to generate decisions and explanations for high-stakes financial tasks, but these explanations can contain factual inconsistencies with the underlying claim or invalid causal reasoning. We propose a two-step auditing framework that decomposes each explanation into reasoning points and verifies them using natural language inference (NLI): (i) claim $\Rightarrow$ point for factuality verification and (ii) point $\Rightarrow$ conclusion for causality verification. To reduce reliance on expensive expert annotations, we study effective fine-tuning strategies based on semi-supervised learning (SSL) for NLI verifiers, leveraging abundant unlabeled explanations via pseudo-labeling and FreeMatch with self-adaptive confidence thresholds. Experiments across small language models and both encoder-only and decoder-only verifiers under varying labeling budgets show that SSL consistently outperforms the pre-trained baseline and supervised fine-tuning. FreeMatch is generally more robust than simple pseudo-labeling, suggesting better utilization of unlabeled data for scaling error detection of LMs' generations."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="liu-etal-2026-semi">
<titleInfo>
<title>Semi-Supervised Learning to Improve Detection of Factual and Causal Errors in Language Models’ Explanations</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mingxuan</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Han</namePart>
<namePart type="family">Yuan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yilin</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Li</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zheng</namePart>
<namePart type="family">Ma</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>Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently (GroundLM 2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yimu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yee</namePart>
<namePart type="given">Man</namePart>
<namePart type="family">Choi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Di</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Siqi</namePart>
<namePart type="family">Ouyang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mozhgan</namePart>
<namePart type="given">Nasr</namePart>
<namePart type="family">Azadani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yi</namePart>
<namePart type="given">R</namePart>
<namePart type="given">(May)</namePart>
<namePart type="family">Fung</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">979-8-89176-478-1</identifier>
</relatedItem>
<abstract>Language models (LMs) are increasingly used to generate decisions and explanations for high-stakes financial tasks, but these explanations can contain factual inconsistencies with the underlying claim or invalid causal reasoning. We propose a two-step auditing framework that decomposes each explanation into reasoning points and verifies them using natural language inference (NLI): (i) claim \Rightarrow point for factuality verification and (ii) point \Rightarrow conclusion for causality verification. To reduce reliance on expensive expert annotations, we study effective fine-tuning strategies based on semi-supervised learning (SSL) for NLI verifiers, leveraging abundant unlabeled explanations via pseudo-labeling and FreeMatch with self-adaptive confidence thresholds. Experiments across small language models and both encoder-only and decoder-only verifiers under varying labeling budgets show that SSL consistently outperforms the pre-trained baseline and supervised fine-tuning. FreeMatch is generally more robust than simple pseudo-labeling, suggesting better utilization of unlabeled data for scaling error detection of LMs’ generations.</abstract>
<identifier type="citekey">liu-etal-2026-semi</identifier>
<location>
<url>https://aclanthology.org/2026.groundlm-1.11/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>141</start>
<end>147</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Semi-Supervised Learning to Improve Detection of Factual and Causal Errors in Language Models’ Explanations
%A Liu, Mingxuan
%A Yuan, Han
%A Wu, Yilin
%A Zhang, Li
%A Ma, Zheng
%Y Wang, Yimu
%Y Choi, Yee Man
%Y Wu, Di
%Y Ouyang, Siqi
%Y Azadani, Mozhgan Nasr
%Y Fung, Yi R. (May)
%S Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently (GroundLM 2026)
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-478-1
%F liu-etal-2026-semi
%X Language models (LMs) are increasingly used to generate decisions and explanations for high-stakes financial tasks, but these explanations can contain factual inconsistencies with the underlying claim or invalid causal reasoning. We propose a two-step auditing framework that decomposes each explanation into reasoning points and verifies them using natural language inference (NLI): (i) claim \Rightarrow point for factuality verification and (ii) point \Rightarrow conclusion for causality verification. To reduce reliance on expensive expert annotations, we study effective fine-tuning strategies based on semi-supervised learning (SSL) for NLI verifiers, leveraging abundant unlabeled explanations via pseudo-labeling and FreeMatch with self-adaptive confidence thresholds. Experiments across small language models and both encoder-only and decoder-only verifiers under varying labeling budgets show that SSL consistently outperforms the pre-trained baseline and supervised fine-tuning. FreeMatch is generally more robust than simple pseudo-labeling, suggesting better utilization of unlabeled data for scaling error detection of LMs’ generations.
%U https://aclanthology.org/2026.groundlm-1.11/
%P 141-147
Markdown (Informal)
[Semi-Supervised Learning to Improve Detection of Factual and Causal Errors in Language Models’ Explanations](https://aclanthology.org/2026.groundlm-1.11/) (Liu et al., GroundLM 2026)
ACL