Search papers, labs, and topics across Lattice.
University of Southern California
3
0
6
Voice input significantly hampers LLM accuracy due to structural transcription issues, while keyboard input is surprisingly more resilient.
Calibration quality of VLMs varies dramatically, with some models exhibiting near-perfect confidence estimates while others are severely overconfident, impacting automation decisions in document processing.
FlowBank reveals that a compact, adaptive portfolio of workflows can outperform traditional single-query generation methods, enhancing efficiency and effectiveness in multi-agent systems.
