Yuchen Jin· @Yuchenj_UW · X·· 1 小时前AI 评分50
AI 导读
MIT CSAIL 等团队提出 Weight-Space Programming，让 GPT-6 Astra 直接读写神经网络权重。仅改动 Qwen3.5-0.8B 的 0.016% 权重，无需梯度或训练，MATH-500 从 7.9% 升至 48.1%，指令跟随提升 3 倍，并修复了官方模型卡警告的推理循环问题。博客：https://rikiyat.github.io/blog/weight-space-programming/
正文 · 原文
This is cool as heck!
Imagine using a large LLM to directly edit a smaller model's weights and dramatically improve its capabilities.
It's like gene editing for neural networks. Even cooler than distillation!
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math. It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training. We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵 Blog: https://rikiyat.github.io/blog/weight-space-programming/在 X 查看被引用的帖子
来源：Yuchen Jin · x.com