Search papers, labs, and topics across Lattice.
Griffith University
1
0
2
Merging slow and fast-thinking models can cut reasoning verbosity by over 24% without sacrificing accuracy, revolutionizing LLM-based recommendations.
