Search papers, labs, and topics across Lattice.
B (Bai et al., 2025a; Team, 2026). We use approximately
3
0
5
Token-level contrastive signals can streamline reasoning processes, leading to more efficient AI models without sacrificing performance.
Rationale-guided knowledge distillation boosts stance detection accuracy in low-resource languages by effectively transferring reasoning knowledge from powerful models to compact student models.
SAYRE's innovative approach to synthesizing KIE training data leads to substantial performance gains for on-device models, particularly in challenging extraction scenarios.
