Search papers, labs, and topics across Lattice.
2
3
4
3
LRMs can be made both more accurate AND more efficient, without training, by dynamically steering their reasoning trajectories based on real-time confidence metrics.
CLIP's image tokens struggle to aggregate information from spatially or semantically related regions, but DeCLIP fixes this by decoupling self-attention and distilling knowledge from VFMs and diffusion models.
