Search papers, labs, and topics across Lattice.
Affiliation:
5
0
8
2
VLMs exhibit a significant performance gap in visual text understanding, with even the best models falling short of human accuracy in error correction tasks.
Residual accessibility in LLMs reveals a surprising correlation with recovery speed, challenging the notion that optimizing for audit scores leads to genuine knowledge deletion.
LLM judges are swayed by style, mistaking it for substance, but a new framework can significantly mitigate this bias.
A novel zero-parameter gating mechanism boosts UAV video segmentation performance by over 4% while maintaining temporal stability.
Even the best vision-language models struggle to reliably set fine-grained GUI states, achieving only 33% accuracy on a new benchmark, but targeted visual hints suggest a clear path to improvement.
