paper-with-me

홈 › Papers

Edu-MMBias: A Three-Tier Multimodal Benchmark for Auditing Social Bias in Vision-Language Models under Educational Contexts

2026-04-11 · Ruijia Li, Mingzi Zhang, Zengyi Yu, Yuang Wei, Bo Jiang arxiv

As Vision-Language Models (VLMs) become integral to educational decision-making, ensuring their fairness is paramount. However, current text-centric evaluations neglect the visual modality, leaving an unregulated channel for latent social biases. To bridge this gap, we present Edu-MMBias, a systematic auditing framework grounded in the tri-component model of attitudes from social psychology. This framework diagnoses bias across three hierarchical dimensions: cognitive, affective, and behavioral. Utilizing a specialized generative pipeline that incorporates a self-correct mechanism and human-in-the-loop verification, we synthesize contamination-resistant student profiles to conduct a holistic stress test on state-of-the-art VLMs. Our extensive audit reveals critical, counter-intuitive patterns: models exhibit a compensatory class bias favoring lower-status narratives while simultaneously harboring deep-seated health and racial stereotypes. Crucially, we find that visual inputs act as a safety backdoor, triggering a resurgence of biases that bypass text-based alignment safeguards and revealing a systematic misalignment between latent cognition and final decision-making. The contributions of this paper are available at: https://anonymous.4open.science/r/EduMMBias-63B2.

📄 PDF Abstract BibTeX arXiv:2604.10200

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

MultiModal Bias: Introducing a Framework for Stereotypical Bias Assessment beyond Gender and Race in Vision Language Models

2023-03-16 · Sepehr Janghorbani, Gerard de Melo

Recent breakthroughs in self supervised training have led to a new class of pretrained vision language models. While there have been investigations of bias in multimodal models, they have mostly focused on gender and rac…

QUACK: Questioning, Understanding, and Auditing Communicated Knowledge in Multimodal Social Deduction Agents

2026-05-26 · Ye Yuan, Rui Song, Weien Li, Zeyu Li 외 arxiv

Social deduction games have become a popular testbed for probing reasoning, deception, coordination, and belief modeling in Large Language Model (LLM) agents. However, most environments are scored only by game outcomes s…

MedBench v5: A Dynamic, Process-Oriented, and Hallucination-Aware Benchmark for Clinical Multimodal Models

2026-06-23 · Jinru Ding, Chuchu Jiang, Lu Lu, Wenrao Pang 외 arxiv

Existing medical AI benchmarks lack process visibility, atomic skill evaluation, and integrated hallucination detection. We introduce MedBench v5, a redesigned benchmark for clinical multimodal models (language, vision-l…

BenchGuard: Who Guards the Benchmarks? Automated Auditing of LLM Agent Benchmarks

2026-04-27 · Xinming Tu, Tianze Wang, Yingzhou, Lu 외 arxiv

As benchmarks grow in complexity, many apparent agent failures are not failures of the agent at all - they are failures of the benchmark itself: broken specifications, implicit assumptions, and rigid evaluation scripts t…

Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack

2026-05-12 · Hao Wang, Hanchen Li, Qiuyang Mang, Alvin Cheung 외 arxiv

Agent benchmarks have become the de facto measure of frontier AI competence, guiding model selection, investment, and deployment. However, reward hacking, where agents maximize a score without performing the intended tas…