Skip to content
AI News HubLIVE
Original source2 min read

Diagnosing and Correcting Bias in MLLM for Long Video Understanding

Summary

The paper argues that long-video question answering is limited less by frame sparsity alone than by bias from spurious linguistic correlations and salient but irrelevant visual cues. It introduces the S-MME diagnostic benchmark and CLEAR, a training-free framework that estimates bias through augmented views and intervenes on hidden states at inference, showing consistent gains across five benchmarks.

SourcearXiv Computer VisionAuthor: Xusheng Liang, Jianqiao Sun, Hao Zhang, Yulei Niu, Hengshuang Zhao, Jiawei Ma
Diagnosing and Correcting Bias in MLLM for Long Video Understanding
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 26 Oct 2026]

Title:Diagnosing and Correcting Bias in MLLM for Long Video Understanding

View a PDF of the paper titled Diagnosing and Correcting Bias in MLLM for Long Video Understanding, by Xusheng Liang and 5 other authors

View PDF HTML (experimental)

Abstract:Long-video question answering is challenging since the answer often hinges on a few decisive moments scattered throughout the video, while memory constraints force Multimodal Large Language Models (MLLMs) to sample frames at extremely low rates, resulting in sparsity that risks omitting key evidence. Though aggressively enlarging the context size is intuitive, we argue that performance remains fundamentally constrained by the bias, \emph{i.e.}, the misleading cues caused by spurious \textit{linguistic} correlations and salient yet irrelevant \textit{visual} observations. In this paper, we first introduce {S-MME}, a benchmark designed to systematically diagnose shortcut-biased behavior in MLLMs. To mitigate this failure, we further propose a training-free framework, Counterfactual Long-video Evidence-Aware Reasoning (CLEAR). By treating the full video-question pair as a complete view, CLEAR estimates bias effects via augmented views by only keeping the question or locally salient visual clips. Then, we intervene hidden states to mitigate the discrepancy at inference such that the correct answer can be predicted confidently. With experimental studies across five benchmarks, we show that our method is generalizable and can be integrated in different MLLMs, yielding consistent gains in prediction and robustness. We hope CLEAR can offer a complementary direction in test-time scaling for long-video understanding.

Comments: 10 pages, 6 figures. Code:this https URL

Subjects:

Computer Vision and Pattern Recognition (cs.CV)

Cite as: arXiv:2610.36210 [cs.CV]

(or arXiv:2610.36210v1 [cs.CV] for this version)

https://doi.org/10.48550/arXiv.2610.36210

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Xusheng Liang [view email] [v1] Mon, 26 Oct 2026 19:44:17 UTC (3,888 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled Diagnosing and Correcting Bias in MLLM for Long Video Understanding, by Xusheng Liang and 5 other authors

View PDF

HTML (experimental)

TeX Source

view license

Additional Features

Audio Summary

Current browse context:

cs.CV

new | recent | 2026-10

Change to browse by:

cs

References & Citations

SciX

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • Introduces S-MME, a benchmark for systematically diagnosing shortcut-biased behavior in MLLMs on long-video tasks.
  • Proposes CLEAR, a training-free framework that estimates bias effects using augmented views keeping only the question or locally salient clips.
  • Intervenes on hidden states at inference to reduce the discrepancy between full and augmented views, enabling more confident correct predictions.
  • Experiments across five benchmarks show the method is generalizable across MLLMs and yields consistent prediction and robustness gains.

Highlights and analysis are generated automatically and may contain errors. Check the original source.

More recent stories on this topic · Models