Search papers, labs, and topics across Lattice.
4
0
8
VLMs still can't reason about spatial logic in real-world scenes, but a new benchmark and scene graph method shows how to make progress.
LLMs can turn sparse rewards into dense training signals for RL agents, achieving comparable performance with significantly higher sample efficiency.
LLMs can be coaxed into generating reliable and economical actions for complex infrastructure management, like power grids, by combining constrained action grammars with preference-based learning.
Forget RLHF, a new framework distills clinician preferences into reusable "HealthPrinciples" that let smaller models outperform giants on medical benchmarks.
