Yu He
PulseAugur coverage of Yu He — every cluster mentioning Yu He across labs, papers, and developer communities, ranked by signal.
-
New ERPO method stabilizes LLM training by controlling query distribution
Researchers have introduced Environment-Regularized Policy Optimization (ERPO), a new method for optimizing Large Language Models (LLMs) that addresses the stability-exploration dilemma. ERPO shifts regularization from …
-
Schrödinger's Navigator enhances robot zero-shot navigation in complex environments
Researchers have developed a new framework called Schrödinger's Navigator for zero-shot object navigation in robots. This system addresses challenges in real-world environments, such as occlusion and unseen hazards, by …
-
New framework decomposes real-world images into layers
Researchers have developed a new framework for decomposing real-world images into layers, addressing limitations in current generative models that are primarily effective in graphic design. Their approach includes an Ag…