Search papers, labs, and topics across Lattice.
3
0
5
3
A simple scoring signal is insufficient for self-distillation; explicitly generating a critique before scoring unlocks corrective credit and enables models to solve previously intractable reasoning problems.
Ditch the expensive PET scan: this method predicts Alzheimer's amyloid-beta status directly from MRI with surprisingly good accuracy.
LLMs struggle to solve topological grid puzzles not because they can't reason, but because they can't reliably extract and maintain spatial constraints.
