Search papers, labs, and topics across Lattice.
Johns Hopkins University
4
0
8
Nexus Sampling retains crucial tokens during KV cache eviction, achieving near-dense attention performance with dramatically reduced memory usage.
MT-EditFlow bridges the gap between local planning and global success in multi-turn image editing, achieving a significant performance boost over leading models.
Agents struggle to match human performance in long-horizon tasks, revealing critical gaps in their learning capabilities during deployment.
A clever two-stage agent using smaller models can produce better, more substantive peer reviews than brute-force application of the largest LLMs.
