Search papers, labs, and topics across Lattice.
1
0
2
LLMs can iteratively refine their reasoning and reduce errors by recursively evaluating and improving their own confidence, leading to more stable and faster inference.
