AI model peers inside and improves other neural networks
🤖 AI-generated content — The title and summary were produced automatically by artificial intelligence, without human editorial review.
MIT researchers showed that GPT-6 Astra can interpret raw internal values in other AI models, predict the functions of different components and edit the systems to improve them, including boosting mathematical abilities. The findings appeared in a preprint on 2 November.
Continue in the app — vote & join in ➔Source: Nature News · via ahirlevel.hu