الباحثون

Shuo Lu

المنشورات 2

نسخة أولية وصول مفتوح

WorldSolver: Can LLM Agents Simulate the Physical Dynamics via Solver Generation?

Siru Jiang, Yongzhe Lyu, Shuo Lu وآخرون · 2026

LLM-based agents are increasingly advancing scientific and engineering problem solving, with physics simulation emerging as a challenging yet practical testbed for reproducing complex physical phenomena with application in embodied AI, games and films. As the workhorse of such simulation, a solver computes how the stat …

نسخة أولية وصول مفتوح

HarnessPAI: An Evolving Harness for Physical AI

Xin Wang, Wenhao Wu, Menghao Zhang وآخرون · 2026

Physical AI aims to build embodied agents that perceive the world, understand and reason about it, and decide how to act. Yet the field has focused primarily on the last component: the action model that maps observations to low-level controls. The prevailing training recipe can erode the perceptual and reasoning capabi …

المؤلفون المشاركون