Search papers, labs, and topics across Lattice.
1
0
3
Forget attention weights – token activation strength in BERT's Layer 8 directly reveals semantically salient words, offering a simpler path to explainability.