Search papers, labs, and topics across Lattice.
7
0
9
0
Super-resolved atmospheric data can now be trusted to respect fundamental physics, enhancing extreme event detection accuracy.
Pruning VLA models can reduce parameters by up to 30% while retaining 90% of performance, challenging the notion of parameter redundancy in these complex systems.
By intelligently pruning attention heads based on their spatial or temporal roles and adaptively routing denoising steps through the network, PARE achieves significant computational savings in video generation without sacrificing quality.
Forget rigid shot-music mappings: BEAT's elastic alignment framework finally captures the dynamic rhythm of professional movie trailer editing.
Autonomous research agents need to learn from their mistakes and adapt, not just generate papers, and this framework shows how to make that happen.
Task-aware localization, using attention cues from both source and target image streams, significantly reduces over-editing in instruction-based image editing, even when applied to strong diffusion transformer backbones.
MLLMs can achieve 4x faster inference without sacrificing accuracy by intelligently focusing on only the image regions relevant to the query.
