Search papers, labs, and topics across Lattice.
Heinrich Hertz Institute Berlin
1
0
3
Large language models may ace rare medical event classification, but their uncertainty estimates are surprisingly unreliable, especially compared to smaller, fine-tuned decoders.