Siheng Chen
Siheng Chen is a tenure-track Associate Professor at Shanghai Jiao Tong University's School of AI and a Principal Investigator at Shanghai Artificial Intelligence Laboratory. He holds a PhD in Electrical and Computer Engineering from Carnegie Mellon University and previously worked as an autonomy engineer at Uber's Advanced Technologies Group and as a research scientist at Mitsubishi Electric Research Laboratories. Chen is best known for foundational contributions to graph signal processing and graph neural networks, for which he received the 2018 IEEE Signal Processing Society Young Author Best Paper Award, and more recently for his work on embodied AI including the TAPT framework for aligning Vision-Language-Action models with tool-based task decomposition.
“TAPT constructs subtask-centric data by pairing bounded subtasks with precise language instructions, so that each VLA tool learns a clear correspondence between the agent's invocation and the intended physical behavior”
Source→“The VLA predicts progress with an auxiliary head p̂_t = ψ_ω(b_t) attached to the backbone feature b_t”
Source→“Siheng Chen, SJTU, corresponding senior author... Previously published in signal processing and graph neural networks; increasingly focused on embodied AI and robot learning.”
Source→AI-extracted from podcast / newsletter / paper summaries. May contain errors.