Search papers, labs, and topics across Lattice.
Fudan University
1
0
3
Current LLMs struggle with writing tasks because existing reward models are too coarse, but a new fine-grained reward modeling approach closes the gap.
