Search papers, labs, and topics across Lattice.
MBZUAI
4
0
10
5
Current egocentric video generation models fail to execute multi-step manipulation tasks realistically, often skipping crucial actions and exhibiting poor physical dynamics.
Fine-tuning VLMs for regional relevance doesn't have to sacrifice global performance: a simple data filtering and model merging technique boosts cultural relevance by 5-15% while barely impacting overall accuracy.
VLMs struggle with basic counting not because they can't "see" the objects, but because they forget to look when generating the answer.
VLMs can regain lost linguistic prowess without extra parameters or architectural changes, thanks to a clever KV-cache sharing trick for distillation.
