Search papers, labs, and topics across Lattice.
Affiliation:
1
0
2
Fine-tuning small-to-medium LLMs can be dramatically improved with just two online rollouts, reshaping efficiency in model training.
