INFACT: A Diagnostic Benchmark for Induced Faithfulness and Factuality Hallucinations in Video-LLMs
Yangjunqi, Yuecong Min, Jie Zhang Institute of Computing Technology, Chinese Academy of Sciences, Shiguang Shan Institute of Computing Technology, Chinese Academy of Sciences, Xilin Chen
Despite rapid progress, Video Large Language Models (Video-LLMs) remain unreliable due to hallucinations, which are outputs that contradict either video evidence (faithfulness) or verifiable world knowledge (factuality).Existing benchmarks provide limited coverage of factuality hallucinations and predominantly evaluate models only in clean settings.We introduce INFACT, a diagnostic benchmark comprising 9,800 QA instances with fine-grained taxonomies for faithfulness and factuality, spanning real and synthetic videos.INFACT evaluates models in four modes: Base (clean), Visual Degradation, Evidence Corruption, and Temporal Intervention for order-sensitive items.Reliability under induced modes is quantified using Resist Rate (RR) and Temporal Sensitivity Score (TSS).Experiments on 14 representative Video-LLMs reveal that higher Base-mode accuracy does not reliably translate to higher reliability in the induced modes, with evidence corruption reducing stability and temporal intervention yielding the largest degradation.Notably, many open-source baselines exhibit near-zero TSS on factuality, indicating pronounced temporal inertia on order-sensitive questions.
尚无 AI 解读，展示原文摘要
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表