Metraipolis › archive › dc_ade9eff1c4fc
From Forecasting to Feedback: Evaluating Learned Dynamics for Tokamak Control — arXiv:2611.03690
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.03690
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-04T12:28:51.802Z
- SHA-256 of the archived text
1e95685aad0324e80007e3830c3f421523513476b70c460b14de36aebf4f4c94- Stable URL of this page
https://metraipolis.com/documento/dc_ade9eff1c4fc
How to cite
From Forecasting to Feedback: Evaluating Learned Dynamics for Tokamak Control — arXiv:2611.03690 — arXiv (en), https://arxiv.org/abs/2611.03690 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ade9eff1c4fc (Metraipolis archive dc_ade9eff1c4fc, sha256 1e95685aad0324e80007e3830c3f421523513476b70c460b14de36aebf4f4c94)
BibTeX
@misc{dcade9eff1c4fc,
title = {From Forecasting to Feedback: Evaluating Learned Dynamics for Tokamak Control — arXiv:2611.03690},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.03690},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ade9eff1c4fc; sha256 1e95685aad0324e80007e3830c3f421523513476b70c460b14de36aebf4f4c94},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03690. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Learned temporal models are moving beyond offline forecasting toward interactive environments for decision making, yet logged-trajectory accuracy need not predict closed-loop behavior under policies optimized in the learned model. This forecasting-control gap matters in tokamak magnetic control, where coupled plasma-circuit dynamics and actuator constraints can amplify errors through feedback. Using a calibrated free-boundary Grad-Shafranov solver, we collect four datasets in a 2x2 design over physics-parameter and current-reference coverage. At matched budgets of 500, 1k, 3k, and 5k rollouts, we train 16 action-conditioned Transformers; PPO policies optimized in the 5k surrogates are then evaluated without adaptation on six reference-solver conditions. Only joint-coverage D4 shows monotonically decreasing cross-condition mean error, whereas D1 improves in distribution but degrades out of distribution. Binary completion masks substantial closed-loop differences: D2-D4 each complete 83.3% of conditions, but reach on average 99.0%, 94.7%, and 90.5% of the evaluation horizon, while their worst conditions reach 94.2%, 68.2%, and 43.2%. Policies trained with action noise complete 5/24 matched conditions, compared with 19/24 under nominal training, and have a lower mean horizon-capped episode length. Based on one trained policy per configuration and one evaluation realization per condition, the results support combining prediction error, episode completion, condition-level duration, and tracking quality. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Plasma Physics (physics.plasm-ph) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2611.03690 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-04T12:26:47.208Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).