Skip to contentDEEKOYA AI CoursesFLAT 20% OFFEnroll Now
Deekoya AI

GPT-6 Astra sees the matrix and reprograms other AI models directly through their weights

The Decoder ·

Summary: MIT researchers show that OpenAI's GPT-6 Astra can read and deliberately rewrite the weights of other AI models without traditional training. In one experiment, Astra boosted a Qwen model's math accuracy from 7.9 to 48.1 percent by changing just 0.016 percent of its weights. The

Read the full story on The Decoder

Explore more: Today in AI · Research papers · AI tools · Learn AI

Related stories

Who reported “GPT-6 Astra sees the matrix and reprograms other AI models directly through…”?

The Decoder, on 4 Nov 2026. The full article is at the source link on this page.