@inproceedings{feng-rudzicz-2026-causalprism,
title = "{C}ausal{P}rism: Component-Level Evaluation of Intervention-Based Causal Reasoning in Large Language Models",
author = "Feng, Jinyue and
Rudzicz, Frank",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1294/",
pages = "27364--27389",
ISBN = "979-8-89176-462-0",
abstract = "Evaluating causal reasoning in large language models (LLMs) is complicated by benchmarks that conflate distinct reasoning operations into aggregate scores. We propose CausalPrism, an evaluation framework grounded in structural causal model semantics and cognitive models of causal inference, which decomposes intervention-based causal reasoning into three functional primitives{---}Causal Prior, Forward State Tracing, and Belief Updating{---}plus a consistency primitive, Agency Sensitivity. Primitives are instantiated in directed acyclic graphs and evaluated across three narrative domains of varying semantic richness. Across 14 open-source and frontier LLMs, Belief Updating is the dominant bottleneck: the best model reaches only 0.613 accuracy, while GPT-5.2 and GPT-5-mini score 0.531 and 0.517 despite near-ceiling performance on Causal Prior and Forward State Tracing. Prior exhibits systematic domain sensitivity and over-rejection. Agency framing effects are family-dependent for weaker models. We argue against treating causal reasoning as a unitary capability and motivate targeted evaluation and training along multiple primitive axes."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="feng-rudzicz-2026-causalprism">
<titleInfo>
<title>CausalPrism: Component-Level Evaluation of Intervention-Based Causal Reasoning in Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jinyue</namePart>
<namePart type="family">Feng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Frank</namePart>
<namePart type="family">Rudzicz</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>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Evaluating causal reasoning in large language models (LLMs) is complicated by benchmarks that conflate distinct reasoning operations into aggregate scores. We propose CausalPrism, an evaluation framework grounded in structural causal model semantics and cognitive models of causal inference, which decomposes intervention-based causal reasoning into three functional primitives—Causal Prior, Forward State Tracing, and Belief Updating—plus a consistency primitive, Agency Sensitivity. Primitives are instantiated in directed acyclic graphs and evaluated across three narrative domains of varying semantic richness. Across 14 open-source and frontier LLMs, Belief Updating is the dominant bottleneck: the best model reaches only 0.613 accuracy, while GPT-5.2 and GPT-5-mini score 0.531 and 0.517 despite near-ceiling performance on Causal Prior and Forward State Tracing. Prior exhibits systematic domain sensitivity and over-rejection. Agency framing effects are family-dependent for weaker models. We argue against treating causal reasoning as a unitary capability and motivate targeted evaluation and training along multiple primitive axes.</abstract>
<identifier type="citekey">feng-rudzicz-2026-causalprism</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1294/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>27364</start>
<end>27389</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CausalPrism: Component-Level Evaluation of Intervention-Based Causal Reasoning in Large Language Models
%A Feng, Jinyue
%A Rudzicz, Frank
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F feng-rudzicz-2026-causalprism
%X Evaluating causal reasoning in large language models (LLMs) is complicated by benchmarks that conflate distinct reasoning operations into aggregate scores. We propose CausalPrism, an evaluation framework grounded in structural causal model semantics and cognitive models of causal inference, which decomposes intervention-based causal reasoning into three functional primitives—Causal Prior, Forward State Tracing, and Belief Updating—plus a consistency primitive, Agency Sensitivity. Primitives are instantiated in directed acyclic graphs and evaluated across three narrative domains of varying semantic richness. Across 14 open-source and frontier LLMs, Belief Updating is the dominant bottleneck: the best model reaches only 0.613 accuracy, while GPT-5.2 and GPT-5-mini score 0.531 and 0.517 despite near-ceiling performance on Causal Prior and Forward State Tracing. Prior exhibits systematic domain sensitivity and over-rejection. Agency framing effects are family-dependent for weaker models. We argue against treating causal reasoning as a unitary capability and motivate targeted evaluation and training along multiple primitive axes.
%U https://aclanthology.org/2026.findings-emnlp.1294/
%P 27364-27389
Markdown (Informal)
[CausalPrism: Component-Level Evaluation of Intervention-Based Causal Reasoning in Large Language Models](https://aclanthology.org/2026.findings-emnlp.1294/) (Feng & Rudzicz, Findings 2026)
ACL