Search papers, labs, and topics across Lattice.
2
0
2
LLMs may struggle in cold-start scenarios, but a novel learned hybrid fusion layer can recover significant retrieval performance, outperforming traditional methods across multiple domains.
Calibrated retrieval-budget allocation can dramatically enhance retrieval-augmented generation by intelligently deciding when to fetch context, leading to more efficient and accurate responses.
