6 papers
From Perception to Prediction and Intervention: Benchmarking Proactive Video Assistants in Real-Life Scenarios
Boyu Cui, Junhao Luo, Zhihao Wang +5
Recent multimodal large language models have advanced video understanding, but existing benchmarks mainly evaluate whether models can interpret observed visual content. In real-wor…
Cooldown Landmines: Cross-Tenant Interference Attacks on LLM Gateways
Yudong Gao, Linghan Chen, Wenhan Wu +8
LLM gateways enforce separate tenant quotas while sharing model deployments and cooldown records that temporarily exclude failing backends. However, a tenant's request failure can…
Bit-Flip Attacks on Vision-Language-Action Models: Action-Decoding Architecture Shapes the Vulnerability
Yudong Gao, Linghan Chen, Wenhan Wu +5
Quantized Vision-Language-Action (VLA) models expose a weight-fault surface: Rowhammer-style faults can corrupt deployed INT8 bits. We present the first bit-flip attack on a VLA: a…
Shortcut Before Circuit: Document Statistics Time In-Context Conflict Resolution
Yijun Liao, Fanwei Liang
When a context asserts two values for one fact, a model commits to a cue -- recency, repetition, position -- but natural data rarely makes these disagree, so behavior cannot reveal…
PathMark: Protecting Intellectual Property of Mixture-of-Expert LLMs via Path Watermarks
Yudong Gao, Qingyue Wang, Yuanyuan Yuan +4
Mixture-of-Experts (MoE) large language models represent high-value intellectual property, yet existing watermarking schemes designed for dense models fail on MoE architectures due…
Attacking the Trusted Imagination: Oracle-Level Integrity Attacks on Imagine-then-Act World Models
Linghan Chen, Kaiyan Ji, Minyu Guo
Many recent vision-language-action (VLA) policies adopt an imagine-then-act design. A world-action model (WAM) first imagines a short future as a latent trajectory z~, on which the…