{"schema":"postcutoff/event@1","as_of":"2026-11-13T13:49:00+01:00","url":"https://postcutoff.com/e/2022-02-16-deepmind-tokamak-plasma-control/","md":"https://postcutoff.com/e/2022-02-16-deepmind-tokamak-plasma-control/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"human editor","policy":"https://postcutoff.com/about/"},"license":null,"id":"2022-02-16-deepmind-tokamak-plasma-control","date":"2022-02-16","date_precision":"day","short_title":"Deep reinforcement learning controls fusion plasma in the TCV tokamak","deck":null,"takeaway":"DeepMind and EPFL (Nature, Feb 2022) trained a single deep-RL policy in simulation that commanded all of TCV's magnetic control coils on the real machine.","category":"science","category_label":"Science & math","importance":4,"confidence":"high","status":{"key":"confirmed","labels":["Result confirmed"]},"sources":[{"n":1,"title":"Magnetic control of tokamak plasmas through deep reinforcement learning (Nature)","url":"https://www.nature.com/articles/s41586-021-04301-9","type":"paper","group":"primary","domain":"nature.com"},{"n":2,"title":"DeepMind: Accelerating fusion science through learned plasma control","url":"https://deepmind.google/blog/accelerating-fusion-science-through-learned-plasma-control/","type":"official","group":"primary","domain":"deepmind.google"}],"official":2,"filed":"2026-09-29","updated":"2026-09-29","orgs":["DeepMind","EPFL Swiss Plasma Center"],"title":"Deep reinforcement learning controls fusion plasma in the TCV tokamak","summary":"DeepMind and EPFL (Nature, Feb 2022) trained a single deep-RL policy in simulation that commanded all of TCV's magnetic control coils on the real machine. It produced and held elongated, negative-triangularity and 'snowflake' plasmas, and even two separate 'droplet' plasmas at once.","key_facts":["Nature 602 (Feb 2022)","Zero-shot sim-to-real transfer: trained in a simulator, deployed directly on the tokamak","One neural controller replaced a set of hand-designed feedback loops for 19 magnetic coils"],"key_numbers":[],"tags":["physics","fusion","reinforcement-learning","control"],"science":{"field":"physics","subfield":"nuclear fusion / plasma control","problem":"Magnetic confinement and shaping of tokamak plasmas","result":"First deep-RL controller to shape and sustain diverse plasma configurations on a real tokamak.","open_since":"","ai_system":["deep reinforcement learning (MPO actor-critic)"],"human_role":"Humans built the simulator, specified targets and rewards, supervised experiments","verification":{"key":"peer-review","label":"Peer-reviewed","text":"Peer-reviewed in Nature; demonstrated on hardware"},"status":"confirmed","shock":""},"body_md":"## What happened\n\nA neural network learned to steer a hot plasma by adjusting magnetic coils thousands of times per second, first in simulation and then on the real reactor.\n\n## Why it matters\n\nIt showed that RL could replace complex hand-engineered control in fusion devices and opened the way to AI-designed plasma scenarios.","disputed":[],"related":[{"id":"2026-09-03-pppl-pacman-fusion-ai-control","url":"https://postcutoff.com/e/2026-09-03-pppl-pacman-fusion-ai-control/","date":"2026-09-03","date_precision":"day","short_title":"PPPL's PACMAN framework lets multiple AI models control a tokamak in ~20 ms, preventing a tearing mode","deck":null,"takeaway":"It is a step toward the AI-supervised operation that future power-plant tokamaks such as SPARC and ITER are expected to need.","category":"science","category_label":"Science & math","importance":2,"confidence":"medium","status":{"key":"confirmed","labels":["Result confirmed"]},"sources":3,"official":1,"filed":"2026-09-29","updated":"2026-09-29","orgs":["Princeton Plasma Physics Laboratory","General Atomics"]},{"id":"2024-02-21-ai-avoids-tokamak-tearing-instabilities","url":"https://postcutoff.com/e/2024-02-21-ai-avoids-tokamak-tearing-instabilities/","date":"2024-02-21","date_precision":"day","short_title":"AI controller predicts and avoids tearing instabilities in the DIII-D fusion reactor","deck":null,"takeaway":"Instabilities that can damage reactors are a key obstacle to fusion power.","category":"science","category_label":"Science & math","importance":3,"confidence":"high","status":{"key":"confirmed","labels":["Result confirmed"]},"sources":2,"official":1,"filed":"2026-09-29","updated":"2026-09-29","orgs":["Princeton University","Princeton Plasma Physics Laboratory","General Atomics"]}],"people":[],"posts":[],"videos":[],"models":[],"changes":[{"date":"2026-09-29","type":"filed","text":"Created"}],"provenance":{"agents":[{"model":"Claude Opus 5.5","maker":"Anthropic","tool":"Claude Code"}],"filed":"2026-09-29","run":null,"sources_read":null,"updated":"2026-09-29","human_review":null,"version":{"date":"2026-10-04"}},"gaps":[{"model_id":"gpt-6-astra","name":"GPT-6 Astra","cutoff":"2026-04","days_after":null,"in_training_data":true},{"model_id":"claude-opus-5-5","name":"Claude Opus 5.5","cutoff":"2026-06","days_after":null,"in_training_data":true},{"model_id":"gemini-4-hydrogen","name":"Gemini 4 Hydrogen","cutoff":"2026-03","days_after":null,"in_training_data":true},{"model_id":"grok-4-7","name":"Grok 4.7","cutoff":"2026-05","days_after":null,"in_training_data":true}],"short_url":"https://postcutoff.com/s/deep-reinforcement-learning"}