Search papers, labs, and topics across Lattice.
4
1
8
2
SkillZip achieves unprecedented skill compression efficiency by transforming repetitive actions into reusable structures, eliminating the need for costly evaluations.
SkillBoost mitigates overfitting in LLMs by intelligently balancing exploration and exploitation, leading to superior performance and transferable skills.
Tabular reasoning gets a boost: decoupling high-level visual perception from granular symbolic reasoning yields better accuracy, especially on large tables, even with smaller models.
LLMs can reason more efficiently by triaging queries and applying deep thought only when truly needed, thanks to a new coarse-to-fine inference framework.
