Search papers, labs, and topics across Lattice.
3
0
6
0
On-demand safety interventions can significantly improve the performance of vision-language models without sacrificing their multimodal reasoning capabilities.
Top LLMs can achieve medal-equivalent scores on elite science exams, but they falter on visual grounding and long-horizon consistency.
SafeCap boosts LVLM safety by up to 19 points through innovative caption-mediated reinforcement learning, outpacing traditional alignment methods.
