Search papers, labs, and topics across Lattice.
1
0
2
7
Scaling language models with interpretability as a core constraint reveals that they can become more capable and understandable simultaneously, defying traditional trade-off assumptions.
