NEWS
 MIT CSAIL: GPT-6 Astra improves open-source model without retrainingMIT CSAIL researchers demonstrated that they improved the mathematical performance of the Qwen3.5-0.8B model by directly editing the weights of GPT-6 Astra. This method corrects structural errors in the model without the need for classical gradient descent. This is a news summary compiled for the AiPulse newsletter — not the original article. The original article is at zdnet.fr. Read the original article AI Critique → English (automatic translation) |