Liang Wang
Liang Wang is a researcher at Microsoft Research whose work focuses on natural language processing and machine learning. He is known for his contributions to text embeddings, including the E5 and multilingual E5 embedding models, as well as research on query expansion with large language models and in-context learning retrieval. His publications have appeared at major conferences such as ACL, EMNLP, SIGIR, and NeurIPS.
“We introduce XEWorld, a controlled cross-embodiment testbed for world models that isolates embodiments by evaluating held-out robots within physically identical scenes”
Source→“We propose StaKe, a plug-in structured supervision framework for VLA fine-tuning that introduces two complementary signals—Stage Supervision (SS) and Keyframe Supervision (KS)—to improve fine-tuning performance while leaving the inference loop unchanged.”
Source→AI-extracted from podcast / newsletter / paper summaries. May contain errors.