Search papers, labs, and topics across Lattice.
2
0
6
8
RAG models struggle to use retrieved knowledge even when it's relevant, but GuarantRAG's two-stage generation and joint decoding boosts accuracy by 12% and slashes hallucinations by 16%.
LRMs can be made more efficient and accurate by strategically adjusting their output length based on task difficulty, leading to a better accuracy-length trade-off.
