Search papers, labs, and topics across Lattice.
3
0
5
11
On-policy reward modeling with LLM judges not only unlocks significant performance gains on complex mathematical reasoning tasks, but also generalizes to improve performance on simpler numerical and multiple-choice benchmarks.
LLMs can now infer plausible stage layouts from unstructured text alone, opening up new possibilities for automated media production.
LLMs can revise scientific papers to significantly improve their predicted citation impact and perceived quality, suggesting a powerful new tool for authors to refine their manuscripts.
