3 accepted papers
Despite recent advances in multimodal reasoning, Multimodal Large Language Models (MLLMs) still struggle on complex tasks where initial visual perceptions can be misleading. This performance gap stems from a critical reasoning flaw we term Visual Inertia: while MLLMs excel at iterative reflection in
Large Reasoning Models (LRMs) have achieved impressive performance on challenging tasks, yet their deep reasoning often incurs substantial computational costs. To achieve efficient reasoning, existing reinforcement learning methods still struggle to construct short reasoning path during the rollout…
The safety mechanisms of large language models (LLMs) exhibit notable fragility, as even fine-tuning on datasets without harmful content may still undermine their safety capabilities. Meanwhile, existing safety alignment methods predominantly rely on the fine-tuning process, which inadvertently lead…