A key capability for video understanding is reliably linking subjects to events across time, yet whether Video Large Language Models (VideoLLMs) actually achieve this remains unclear. In this work, we introduce DistractionBench to evaluate whether VideoLLMs can robustly link subjects and events in the presence of unrelated video segments. Through controlled interventions, such as inserting short advertisement clips into longer videos, we show that VideoLLMs frequently hallucinate interactions between entities from different segments, incorrectly attributing actions from injected advertisements to subjects in the main video. We characterize this systematic hallucination as bag-of-events (BoE) behavior, where models process videos as collections of events rather than temporally structured sequences. Evaluating 11 popular VideoLLMs, we find that all models exhibit substantial BoE behavior. Our findings suggest that VideoLLMs lack reliable mechanisms for temporal grounding and motivate the development of models with more robust subject-event association.
@article{arxiv.2605.27101,
title = {Pop-Up Distractions Reveal Bag-of-Events Behavior in Video Large Language Models},
author = {Oscar Chew and Serhii Honcharenko and Qian-Hui Chen and Patricia Lu and Dishant Zaveri and Khoa D. Doan and Kuan-Hao Huang},
journal= {arXiv preprint arXiv:2605.27101},
year = {2026}
}