paper-with-me

Papers

Spatiotemporal Sycophancy: Negation-Based Gaslighting in Video Large Language Models

2026-04-20 · Ziyao Tang, Pengkun Jiao, Bin Zhu, Huiyan Qi, Jingjing Chen, Yu-Gang Jiang arxiv

Video Large Language Models (Vid-LLMs) have demonstrated remarkable performance in video understanding tasks, yet their robustness under conversational interaction remains largely underexplored. In this paper, we identify spatiotemporal sycophancy, a failure mode in which Vid-LLMs retract initially correct, visually grounded judgments and conform to misleading user feedback under negation-based gaslighting. Rather than merely changing their answers, the models often fabricate unsupported temporal or spatial explanations to justify incorrect revisions. To systematically investigate this phenomenon, we propose a negation-based gaslighting evaluation framework and introduce GasVideo-1000, a curated benchmark designed to probe spatiotemporal sycophancy with clear visual grounding and temporal reasoning requirements. We evaluate a broad range of state-of-the-art open-source and proprietary Vid-LLMs across diverse video understanding tasks. Extensive experiments reveal that vulnerability to negation-based gaslighting is pervasive and severe, even among models with strong baseline performance. While prompt-level grounding constraints can partially mitigate this behavior, they do not reliably prevent hallucinated justifications or belief reversal. Our results indicate that current Vid-LLMs lack robust mechanisms for maintaining grounded spatiotemporal beliefs under adversarial conversational feedback.

📄 PDF Abstract BibTeX arXiv:2604.17873

Code (0)

등록된 구현이 없습니다.

Tasks

Visual Grounding

Similar Papers 제목 키워드 기반

Calling a Spade a Heart: Gaslighting Multimodal Large Language Models via Negation

2025-01-31 · Bin Zhu, Hui yan Qi, Yinxuan Gui, Jingjing Chen 외

Multimodal Large Language Models (MLLMs) have exhibited remarkable advancements in integrating different modalities, excelling in complex understanding and generation tasks. Despite their success, MLLMs remain vulnerable…

Negation

Benchmarking Gaslighting Attacks Against Speech Large Language Models

2025-09-24 · Jinyang Wu, Bin Zhu, Xiandong Zou, Qiquan Zhang 외 arxiv

As Speech Large Language Models (Speech LLMs) become increasingly integrated into voice-based applications, ensuring their robustness against manipulative or adversarial input becomes critical. Although prior work has st…

Can a large language model be a gaslighter?

2024-10-11 · Wei Li, Luyao Zhu, Yang song, Ruixi Lin 외

Large language models (LLMs) have gained human trust due to their capabilities and helpfulness. However, this in turn may allow LLMs to affect users' mindsets by manipulating language. It is termed as gaslighting, a psyc…

Language ModelingLanguage ModellingLarge Language Modelmodel+1

Ask don't tell: Reducing sycophancy in large language models

2026-02-27 · Magda Dubois, Cozmin Ududec, Christopher Summerfield, Lennart Luettgau arxiv

Sycophancy, the tendency of large language models to favour user-affirming responses over critical engagement, has been identified as an alignment failure, particularly in high-stakes advisory and social contexts. While …

Stress Tests REVEAL Fragile Temporal and Visual Grounding in Video-Language Models

2026-02-11 · Sethuraman T, Savya Khosla, Aditi Tiwari, Vidya Ganesh 외 arxiv

This work investigates a fundamental question: Do Video-Language Models (VidLMs) robustly account for video content, temporal sequence, and motion? Our investigation shows that, surprisingly, they often do not. We introd…

Visual Grounding