14 papers
TRACE: Trajectory-Based Safety Patch Learning for LLM Post-Training Realignment
Changyue Li, Jiaming He, Youliang Yuan +4
Fine-Tuning-as-a-Service (FTaaS) platforms let users train large language models (LLMs) on customized tasks, but this pipeline could erode models' safety alignment. In practice, se…
Semantic Router: On the Feasibility of Hijacking MLLMs via a Single Adversarial Perturbation
Changyue Li, Jiaying Li, Youliang Yuan +3
Multimodal Large Language Models (MLLMs) are increasingly deployed in stateless systems, such as autonomous driving and robotics. This paper investigates a novel threat: Semantic-A…
PaSBench-Video: A Streaming Video Benchmark for Proactive Safety Warning
Yusong Zhao, Yuejin Xie, Youliang Yuan +4
Between the first visible sign of danger and the moment an accident occurs, there is often a window where intervention remains possible. Video-capable multimodal large language mod…
Human Cognitive Benchmarks Reveal Foundational Visual Gaps in MLLMs
Jen-Tse Huang, Dasen Dai, Jen-Yuan Huang +7
Humans develop perception through a bottom-up hierarchy: from basic primitives and Gestalt principles to high-level semantics. In contrast, current Multimodal Large Language Models…
SHAPE: Unifying Safety, Helpfulness and Pedagogy for Educational LLMs
Sihang Zhao, Kangrui Yu, Youliang Yuan +2
Large Language Models (LLMs) have been widely explored in educational scenarios. We identify a critical vulnerability in current educational LLMs, pedagogical jailbreaks, where stu…
Curing Miracle Steps in LLM Mathematical Reasoning with Rubric Rewards
Youliang Yuan, Qiuyang Mang, Jingbang Chen +7
In this paper, we observe that current models are susceptible to reward hacking, leading to a substantial overestimation of a model's reasoning ability. This is evidenced by a high…