Search papers, labs, and topics across Lattice.
Affiliation:
3
0
7
Agents can now recursively evolve their own training environments, jointly building and validating tasks, contexts, and verifiers for robust long-horizon learning.
Fusing image features with semantic segmentation and geographic clustering unlocks substantial gains in worldwide image geo-localization, surpassing prior methods by up to 16.58% at the street level.
A 1000x larger video reasoning dataset reveals early signs of emergent generalization, offering a new foundation for training and evaluating spatiotemporal AI.
