Search papers, labs, and topics across Lattice.
Affiliation:
5
0
11
8
Asynchronous LLM training can be made robust to stale rollouts by minimizing a novel group-mass capping bias bound, outperforming existing methods on reasoning tasks.
Visual attention in VLMs is not uniformly reliable; it varies across three distinct modes, revealing a surprising gap between model predictions and human intuition.
Forget pixel-level noise: FogFool shows that physically-plausible, atmospherically-modeled fog can achieve 84% transfer attack success rate against remote sensing image classifiers, even surviving JPEG compression.
