MIT 研究：GPT-6 Astra 可直接读写其他模型权重
原标题：GPT-6 Astra sees the matrix and reprograms other AI models directly through their weights
AI 摘要
MIT 研究人员在博客中展示，OpenAI 的 GPT-6 Astra 能直接读取并修改其他 AI 模型的权重，无需传统训练流程，他们将此称为“权重空间编程”。实验显示 Astra 仅凭原始权重就能识别模型训练过的游戏和语言，还能从零生成游戏模型，并在部分游戏中超过训练基线；在 Qwen3.5-0.8B 上仅改动 0.016% 的权重，就把 MATH-500 准确率从 7.9% 提升到 48.1%。研究者认为该方法可用于权重安全审计和定向行为修正，但能力仍有限，且 Jürgen Schmidhuber 质疑其新颖性。
正文节选
GPT-6 Astra sees the matrix and reprograms other AI models directly through their weights MIT researchers show that GPT-6 Astra can directly understand and modify the internal settings of other AI models without traditional training. The approach has limits but could change how AI systems are tuned and controlled. Hand a person a list of thousands of numbers and ask what game an AI model trained on them plays. You won't get an answer. GPT-6 Astra from OpenAI will give you one: In a research blog