Asterix
An AI industry professional who focuses on open-source model licensing and commercialization compliance issues, with relevant firsthand experience.
Timeline (last 90 days)
In experiment 5, the Breakout, Seaquest, and Asterix networks generated by GPT-6 Astra in a single pass outperformed policy networks trained conventionally for 100 million steps, and on Breakout outperformed a version trained for 500 million steps.
In experiment 5, against a model trained for 500 million steps, Astra's generated networks scored 20 times higher in Breakout, came close in Asterix, and reached 60% of the score in Sequest.
All Facts (2)
In experiment 5, the Breakout, Seaquest, and Asterix networks generated by GPT-6 Astra in a single pass outperformed policy networks trained conventionally for 100 million steps, and on Breakout outperformed a version trained for 500 million steps.
50%UnverifiedIn experiment 5, against a model trained for 500 million steps, Astra's generated networks scored 20 times higher in Breakout, came close in Asterix, and reached 60% of the score in Sequest.
50%