「Weight-Space Programming」の概要をまとめました。 2026年11月2日、MIT CSAILの研究チームは、LLMがニューラルネットワークの重みを直接読み書きする新しい手法「Weight-Space Programming」を発表しました。
SHELDON — Just off the bridge in Sheldon lies a yellow building which the local historical society has taken up residence in and renovated over the past 10 years. The H.H. Mower property at 200 Bridge ...
On November 2, 2026, a research team at MIT CSAIL announced a new method called "Weight-Space Programming", in which an LLM directly reads and writes the weights of a neural network. The "weights" of ...
It’s called “weight space programming”. MIT researchers used Astra to read and edit model weights “directly”. No optimisations, no backprop. Just read and write weights directly using Python. This is ...
The outputs of AI models are the result of a complex interplay between neurons, weights, training, and inference, and are therefore difficult to follow. But in some ways they're starting to look more ...
MIT CSAIL has proposed a new method called "Weight-Space Programming," enabling GPT-6 Astra to directly read and modify neural network weights. In a network upgrade test using the Qwen3.5-0.8B model, ...
The claim comes from Alex Krentsel, an autonomous systems PhD at UC Berkeley and senior researcher at Google Research, who is also known for building Exo, an open-source agent harness designed around ...
The weights of a deep neural network are vast arrays of numbers, ranging from tens of megabytes to hundreds of gigabytes, and humans can hardly read them directly. Weights have traditionally been ...
Four MIT CSAIL researchers had GPT-6 Astra inspect a small Qwen model and patch 131,072 of its 0.8 billion weights with no gradient descent. MATH-500 went from 7.9% to 48.1%. A day later, spaCy ...
Qwen小模型因反覆推理而難以完成作答,MIT研究讓GPT-6 Astra直接找出並修正相關權重,僅調整0.016%的參數,不必重新訓練,就將MATH-500正確率提高至48.1% CYBERSEC 2027 將於 5 月 4 日至 6 日盛大登場 ...
GPT-6 Astra changed two matrices inside Qwen3.5-0.8B and, in the authors’ evaluation, its MATH-500 score rose from 7.9% to 48.1%. GSM8K went from 13.8% to 67.1%, and strict instruction-following on ...
A new MIT CSAIL blog post is making the rounds in AI circles this week, and it describes something that sounds almost like science fiction: a general-purpose AI model reading the raw numerical weights ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results