Search papers, labs, and topics across Lattice.
2
0
6
Why waste tokens on reasoning when you don't need to? Selective Chain-of-Thought cuts LLM inference costs by up to 47% in medical QA, with minimal accuracy loss.
Robot dogs can learn to climb complex U-shaped stairs and generalize to other stair types using a two-stage RL approach, opening doors for more versatile robotic navigation in construction and other environments.
