5 papers
Does Deeper Reasoning Compromise Alignment? Revealing and Mitigating of Alignment Collapse in Large Reasoning Models
Yu-Hang Wu, Yu-Jie Xiong, Henghua Zhang +3
The emergence of Chain-of-Thought (CoT) has established a robust foundation for Large Reasoning Models (LRMs). While deep reasoning is widely believed to enhance safety alignment,…
Freeze Deep, Train Shallow: Interpretable Layer Allocation for Continued Pre-Training
Yu-Hang Wu, Qin-Yuan Liu, Qiu-Yang Zhao +3
Selective layer-wise updates are essential for low-cost continued pre-training of Large Language Models (LLMs), yet determining which layers to freeze or train remains an empirical…
MIND: Monge Inception Distance for Generative Models Evaluation
Quentin Berthet, Yu-Han Wu, Clement Crepy +3
We propose the Monge Inception Distance (MIND), a metric for evaluating generative models that addresses key limitations of the widely adopted Fréchet Inception Distance (FID). The…
Delta Forcing: Trust Region Steering for Interactive Autoregressive Video Generation
Yuheng Wu, Xiangbo Gao, Tianhao Chen +4
Interactive real-time autoregressive video generation is essential for applications such as content creation and world modeling, where visual content must adapt to dynamically evol…
Sugar-Coated Poison: Benign Generation Unlocks LLM Jailbreaking
Yu-Hang Wu, Yu-Jie Xiong, Hao Zhang +2
With the increasingly deep integration of large language models (LLMs) across diverse domains, the effectiveness of their safety mechanisms is encountering severe challenges. Curre…