Hui Chen

Unbiased Token Pruning for Efficient Large Multimodal Models

PruneHal: Reducing Hallucinations in Multi-modal Large Language Models through Adaptive KV Cache Pruning

mmSSR: Harvesting Rich, Scalable and Transferable Multi-Modal Data for Instruction Fine-Tuning

PrefixKV: Adaptive Prefix KV Cache is What Vision Instruction-Following Models Need for Efficient Generation

Towards Efficient Vision-Language Tuning: More Information Density, More Generalizability