Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
Unreasonable Labs
3
0
6
2
LLMs can maintain surprisingly high alignment with ground truth scientific conclusions (F1 score of 0.7) even when given minimal context, suggesting potential for AI-driven scientific discovery.
Looping language models isn't just for single agents anymore: Recursive Multi-Agent Systems (RecursiveMAS) show that agent collaboration itself can be scaled through recursion, yielding faster and more efficient problem-solving.
Verifiable rewards in RL can lead to impressive performance gains in compact LLMs, but don't guarantee robust physical reasoning, instead inducing procedural solution templates.