<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ghermi-etal-2026-cause">
    <titleInfo>
        <title>From Cause to Effect: Generating Intermediate Event Chains in Videos</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Ridouane</namePart>
        <namePart type="family">Ghermi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Vicky</namePart>
        <namePart type="family">Kalogeiton</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Ivan</namePart>
        <namePart type="family">Laptev</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 2026 Conference on Empirical Methods in Natural Language Processing</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">979-8-89176-459-0</identifier>
    </relatedItem>
    <abstract>Video-Language Models (VLMs) excel at visual recognition but struggle with causal reasoning. Current evaluations are flawed: multiple-choice QA allows shortcuts, and open-ended QA lacks robust automatic evaluation. We introduce Causal Chain Generation (CCG), a novel task where models have to generate the sequence of intermediate events connecting a cause and an effect. This structured setup demands genuine reasoning and enables reliable automatic scoring. We accompany this task with two complementary metrics: SODA-Embedding, which measures the semantic matching of a generated chain against a reference, and Causal-QA, which probes the chain’s causal coherence by asking whether it contains the specific causal facts needed to reconstruct the narrative. We provide two test sets: SF20K-CC, a manually annotated diagnostic on 20 short films, and CMD-CC, a movie clip benchmark derived from audio descriptions. Our analysis reveals that state-of-the-art VLMs struggle with narrative causality understanding, frequently hallucinating or omitting key events. To address this, we fine-tune VLMs with GRPO using a composite reward that combines a format reward, SODA-Embedding, and Causal-QA accuracy, demonstrating substantial improvements over strong baselines.</abstract>
    <identifier type="citekey">ghermi-etal-2026-cause</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.2498/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>56724</start>
            <end>56746</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T From Cause to Effect: Generating Intermediate Event Chains in Videos
%A Ghermi, Ridouane
%A Kalogeiton, Vicky
%A Laptev, Ivan
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F ghermi-etal-2026-cause
%X Video-Language Models (VLMs) excel at visual recognition but struggle with causal reasoning. Current evaluations are flawed: multiple-choice QA allows shortcuts, and open-ended QA lacks robust automatic evaluation. We introduce Causal Chain Generation (CCG), a novel task where models have to generate the sequence of intermediate events connecting a cause and an effect. This structured setup demands genuine reasoning and enables reliable automatic scoring. We accompany this task with two complementary metrics: SODA-Embedding, which measures the semantic matching of a generated chain against a reference, and Causal-QA, which probes the chain’s causal coherence by asking whether it contains the specific causal facts needed to reconstruct the narrative. We provide two test sets: SF20K-CC, a manually annotated diagnostic on 20 short films, and CMD-CC, a movie clip benchmark derived from audio descriptions. Our analysis reveals that state-of-the-art VLMs struggle with narrative causality understanding, frequently hallucinating or omitting key events. To address this, we fine-tune VLMs with GRPO using a composite reward that combines a format reward, SODA-Embedding, and Causal-QA accuracy, demonstrating substantial improvements over strong baselines.
%U https://aclanthology.org/2026.emnlp-main.2498/
%P 56724-56746

Markdown (Informal)

[From Cause to Effect: Generating Intermediate Event Chains in Videos](https://aclanthology.org/2026.emnlp-main.2498/) (Ghermi et al., EMNLP 2026)

• From Cause to Effect: Generating Intermediate Event Chains in Videos (Ghermi et al., EMNLP 2026)

ACL

• Ridouane Ghermi, Vicky Kalogeiton, and Ivan Laptev. 2026. From Cause to Effect: Generating Intermediate Event Chains in Videos. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 56724–56746, Budapest, Hungary. Association for Computational Linguistics.