Search papers, labs, and topics across Lattice.
3
0
6
U-TTT adapts dynamically during inference, achieving state-of-the-art PET image denoising even under challenging distribution shifts.
Forget end-to-end training: MedVol-R1 leverages reinforcement learning to decouple evidence grounding from volumetric delineation, achieving state-of-the-art performance in 3D medical image segmentation by grounding clinical reasoning to verifiable 2D evidence.
A new family of GUI agents, GUI-Owl-1.5, leapfrogs existing open-source models on 20+ GUI benchmarks, proving that multi-platform, real-time GUI automation is now within reach.
