@inproceedings{rajesh-etal-2026-data,
title = "Data Augmentation for Improving Contextual Faithfulness Under Knowledge Conflict",
author = "Rajesh, Mohith and
Maheshwari, Raj and
Tiwari, Vashisth and
Bisk, Yonatan and
Strubell, Emma and
Milbauer, Jeremiah and
Fernandez, Jared",
editor = "Zhang, Yue and
Xu, Ruifeng and
Purwarianti, Ayu and
Rong, Tong and
Suzuki, Jun",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {IJCNLP}-{AACL} 2026",
month = nov,
year = "2026",
address = "Hengqin, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-aacl.225/",
pages = "4419--4431",
ISBN = "979-8-89176-444-6",
abstract = "External context enables language models to accomplish tasks which rely on information beyond what is captured during pretraining. A standard way to improve contextual faithfulness in these settings is to fine-tune on synthetic, modified questions for which the model must adhere to provided contexts inconsistent with information that may have been seen during training. While this approach reliably improves contextual faithfulness, we find that adapting to a narrowly targeted counterfactual signal yields degradation on other faithfulness evaluations and regression on tasks that evaluate a model{'}s parametric world knowledge and reasoning. We find that data augmentation, in which a model is trained on \textit{both} parametrically consistent and counterfactual examples improves performance on faithfulness tasks with reduced regression on parametric world knowledge and reasoning tasks. We find that these improvements are largely consistent across post-training objectives (i.e. SFT, DPO) and across ground-truth, counterfactual data augmentation ratios."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="rajesh-etal-2026-data">
<titleInfo>
<title>Data Augmentation for Improving Contextual Faithfulness Under Knowledge Conflict</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mohith</namePart>
<namePart type="family">Rajesh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Raj</namePart>
<namePart type="family">Maheshwari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vashisth</namePart>
<namePart type="family">Tiwari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yonatan</namePart>
<namePart type="family">Bisk</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Emma</namePart>
<namePart type="family">Strubell</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jeremiah</namePart>
<namePart type="family">Milbauer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jared</namePart>
<namePart type="family">Fernandez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruifeng</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ayu</namePart>
<namePart type="family">Purwarianti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tong</namePart>
<namePart type="family">Rong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Suzuki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hengqin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-444-6</identifier>
</relatedItem>
<abstract>External context enables language models to accomplish tasks which rely on information beyond what is captured during pretraining. A standard way to improve contextual faithfulness in these settings is to fine-tune on synthetic, modified questions for which the model must adhere to provided contexts inconsistent with information that may have been seen during training. While this approach reliably improves contextual faithfulness, we find that adapting to a narrowly targeted counterfactual signal yields degradation on other faithfulness evaluations and regression on tasks that evaluate a model’s parametric world knowledge and reasoning. We find that data augmentation, in which a model is trained on both parametrically consistent and counterfactual examples improves performance on faithfulness tasks with reduced regression on parametric world knowledge and reasoning tasks. We find that these improvements are largely consistent across post-training objectives (i.e. SFT, DPO) and across ground-truth, counterfactual data augmentation ratios.</abstract>
<identifier type="citekey">rajesh-etal-2026-data</identifier>
<location>
<url>https://aclanthology.org/2026.findings-aacl.225/</url>
</location>
<part>
<date>2026-11</date>
<extent unit="page">
<start>4419</start>
<end>4431</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Data Augmentation for Improving Contextual Faithfulness Under Knowledge Conflict
%A Rajesh, Mohith
%A Maheshwari, Raj
%A Tiwari, Vashisth
%A Bisk, Yonatan
%A Strubell, Emma
%A Milbauer, Jeremiah
%A Fernandez, Jared
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-444-6
%F rajesh-etal-2026-data
%X External context enables language models to accomplish tasks which rely on information beyond what is captured during pretraining. A standard way to improve contextual faithfulness in these settings is to fine-tune on synthetic, modified questions for which the model must adhere to provided contexts inconsistent with information that may have been seen during training. While this approach reliably improves contextual faithfulness, we find that adapting to a narrowly targeted counterfactual signal yields degradation on other faithfulness evaluations and regression on tasks that evaluate a model’s parametric world knowledge and reasoning. We find that data augmentation, in which a model is trained on both parametrically consistent and counterfactual examples improves performance on faithfulness tasks with reduced regression on parametric world knowledge and reasoning tasks. We find that these improvements are largely consistent across post-training objectives (i.e. SFT, DPO) and across ground-truth, counterfactual data augmentation ratios.
%U https://aclanthology.org/2026.findings-aacl.225/
%P 4419-4431
Markdown (Informal)
[Data Augmentation for Improving Contextual Faithfulness Under Knowledge Conflict](https://aclanthology.org/2026.findings-aacl.225/) (Rajesh et al., Findings 2026)
ACL