Search papers, labs, and topics across Lattice.
Mohamed bin Zayed University of Artificial Intelligence (MBZUAI)
2
0
4
LLMs can be made more reliably factual by adapting uncertainty estimates to the specific prompt, boosting conditional coverage without sacrificing overall accuracy.
Deferring to a larger LLM only when a smaller LLM is uncertain can match the performance of the larger model alone, while slashing inference costs.
