Jiahao Nie

Boosting Knowledge Utilization in Multimodal Large Language Models via Adaptive Logits Fusion and Attention Reallocation

Advancing Cross-domain Discriminability in Continual Learning of Vision-Language Models