Search papers, labs, and topics across Lattice.
Institute of Computing Technology
4
0
8
Search-oriented rubrics redefine how we evaluate document sets, leading to a 2.6-point performance boost in deep research tasks.
Explicitly grounding evidence in spatial relation tasks can boost VLM performance by nearly 12 points, transforming how we approach visual reasoning.
Current reranking methods fall short, but Rubric4Setwise transforms evaluation into actionable selection signals, achieving unprecedented performance across diverse document sets.
By explicitly modeling physical scales within a Transformer architecture, DynFormer slashes PDE solution error by 95% and dramatically reduces memory consumption.
