Search papers, labs, and topics across Lattice.
Affiliation:
2
0
4
Full-solution interaction in multi-agent LLMs can erase diversity, leading to suboptimal performance despite the presence of multiple agents.
OpenAIReview + GPT-5.5 not only predicts paper quality with 83% accuracy but also detects over 71% of critical errors, showcasing the promise of AI in peer review.
