---
id: "2022-02-16-deepmind-tokamak-plasma-control"
url: "https://postcutoff.com/e/2022-02-16-deepmind-tokamak-plasma-control/"
as_of: "2026-11-13T13:49:00+01:00"
date: "2022-02-16"
date_precision: day
category: science
importance: 4
confidence: high
status: [Result confirmed]
verification: Peer-reviewed
sources: 2
editor: human editor
human_review: null
version: "2026-10-04"
---

As of: 2026-11-13 13:49 CET.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/
Canonical page: https://postcutoff.com/e/2022-02-16-deepmind-tokamak-plasma-control/

# Deep reinforcement learning controls fusion plasma in the TCV tokamak

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

## What happened

A 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.

## Why it matters

It showed that RL could replace complex hand-engineered control in fusion devices and opened the way to AI-designed plasma scenarios.

## Your AI and this story

- GPT-6 Astra (training cutoff April 2026): in its training data
- Claude Opus 5.5 (training cutoff June 2026): in its training data
- Gemini 4 Hydrogen (training cutoff March 2026): in its training data
- Grok 4.7 (training cutoff May 2026): in its training data

## Sources

1. [Magnetic control of tokamak plasmas through deep reinforcement learning (Nature)](https://www.nature.com/articles/s41586-021-04301-9) (nature.com, paper)
2. [DeepMind: Accelerating fusion science through learned plasma control](https://deepmind.google/blog/accelerating-fusion-science-through-learned-plasma-control/) (deepmind.google, official)

## Changes

- 2026-09-29 (filed): Created

## Related

- 2026-09-03: [PPPL's PACMAN framework lets multiple AI models control a tokamak in ~20 ms, preventing a tearing mode](https://postcutoff.com/e/2026-09-03-pppl-pacman-fusion-ai-control/index.md)
- 2024-02-21: [AI controller predicts and avoids tearing instabilities in the DIII-D fusion reactor](https://postcutoff.com/e/2024-02-21-ai-avoids-tokamak-tearing-instabilities/index.md)