Search papers, labs, and topics across Lattice.
3
0
8
Tree search on on-policy trajectories dramatically boosts policy gradient learning by uncovering rare high-return paths, outperforming PPO on Atari and MuJoCo with minimal bias increase.
Adaptive layer selection in LLMs can significantly enhance inference efficiency without sacrificing accuracy, achieving better performance than static pruning methods.
LLMs can nail the final answer in code execution but still fail to reason about the steps to get there, exposing a critical flaw in current evaluation methods.
