Search papers, labs, and topics across Lattice.
Affiliation:
2
0
5
0
LLM pruning can be task-aware: DIPrune's dual importance scoring jointly optimizes static feature saliency and dynamic semantic evolution, outperforming prior methods on LLaVA and Qwen-VL.
Stop letting noisy vision-language alignment ruin your referring image segmentation: AML filters out the bad parts.
