Search papers, labs, and topics across Lattice.
3
0
8
0
Forget scaling model size: RefineRL shows that incentivizing self-refinement in smaller LLMs lets them punch *way* above their weight, rivaling models 10x larger on competitive programming tasks.
By cleverly combining YOCO's efficient attention with recursive computation, YOCO-U achieves a capability-efficiency sweet spot that neither technique can reach on its own.
Language models can learn directly from real-world user interactions, boosting performance without human annotations or simulated environments.
