Loading verified research data…

Experiment 5 · MinAtar Breakout

Can a language model write a useful neural network?

We tested whether language models could produce the numerical weights for a small game-playing network directly, without training feedback.

The networks controlled Breakout in MinAtar. Their scores were measured over thousands of episodes, then checked against the preserved experiment files and raw evaluation data.

10 × 10 game board

From text to play

How the experiment works

  1. 1

    Give the task

    The prompt describes Breakout, the network, and the weight format.

  2. 2

    Generate weights

    A model returns numerical values for the network’s connections.

  3. 3

    Run the policy

    The fixed network chooses actions from the board it observes.

  4. 4

    Measure performance

    Thousands of episodes show how many bricks each policy breaks.

Corrected local evaluation

Mean score by condition

Full results →

Research data are loading.

The Astra reference and trained PPO baseline are shown in separate categories from the five zero-shot model generations.

Separate experiment

Argon High — iterative development

Two researcher-directed follow-ups were evaluated as separate versions. Their results are not part of the original five-model zero-shot ranking.

See the iterative results