Search papers, labs, and topics across Lattice.
Fraunhofer Institute for Cognitive Systems IKS Garching, Technical University of Munich
2
0
4
CUBICS reveals that context-aware performance estimation can significantly improve the reliability of safety-critical ML components by avoiding oversimplified failure models.
Multimodal models often exhibit lower confidence than their unimodal counterparts when they're about to fail, and this work leverages that insight to build a better failure detector.