GPT-6 Astra sees the matrix and reprograms other AI models directly through their weights
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…
- 01.MIT researchers show that OpenAI's GPT-6 Astra can read and deliberately rewrite the weights of other AI models without traditional training
- 02.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
This material is provided as an extended analytical summary.
Read full article at sourceRelated Intelligence
All stories →Moonshot AI published the weights of the Kimi K3.1 Pro on Hugging Face. However, to launch locally you will have to fork out some money.
Moonshot AI published the weights of the Kimi K3.1 Pro on Hugging Face. However, to launch locally you will have to fork out some money. The model has 2.8 trillion parameters, and the weights alone reportedly take up about 1.56 TB. To run, you can use a server with eight NVIDIA...
New open-source benchmark reveals reasoning as the biggest bottleneck for AI in cloud operations
CloudOpsBench is a new open-source benchmark that measures how well AI agents handle typical DevOps tasks in the cloud, from fixing services and removing resources to restricting access rights, rolling out updates, and decommissioning projects. The article New open-source…
Uber only fulfilled 4 rides from AI agents last quarter
AI agents won't be stealing Uber's lunch any time soon. Only four rides were arranged by third-party AI agents during the third quarter of 2026, Uber executives disclosed during an earnings call this morning. The volume of Uber Eats orders fulfilled by AI agents was a little…