Reproduce “A Generative Foundation Model for Multimodal Tokamak Tasks” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yingshuai Ji, Haodong Feng, Guohong Deng, Gelin Bian, Yue Wang
GenToka uses a shared schema-conditioned generative model to forecast tokamak signals, infer actuator voltages, and predict disruption risk from heterogeneous measurements. Task schemas specify inputs and requested outputs, generated through conditional diffusion. On MAST, it lowers macro-average NRMSE from 0.2134 to 0.2033 across Groups 1–4; Group 6 F1 rises from 0.5281 to 0.5726. These results do not establish cross-device transfer or closed-loop control.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.