aifollow.news
返回 Rohan Paul
Rohan Paul· @rohanpaul_ai · X· · 原发布时间 AI 评分61

一项研究报告:Qwen3.5-27B通过推理修改自身权重

AI 导读

一则介绍研究的帖子称,研究人员让Qwen3.5-27B在禁用梯度的条件下,利用自己的权重、练习题评分结果和Python会话提出权重修改。在预先确定的490道未见题目测试中,修改后的准确率从57.2%升至63.8%;同一实验中的2B模型没有修改权重,4B和9B模型的修改则导致性能受损。

原帖引用 · 原文

@rohanpaul_ai

– https://t.co/Jqlke4C7nz

Title: "Language Models Can Rewrite Their Own Weights"

回复Rohan Paul@rohanpaul_ai
New Stanford+other top labs paper finds that a 27B open model can improve itself by editing its own weights through reasoning, with no training or gradients. Improving a deployed model today means another training run on large clusters. Prompts and extra inference compute help, but the gain disappears once the task ends. They gave Qwen3.5-27B its own weights, its graded practice answers, and a Python session, with gradients blocked. It compared its hidden states on answers that finished against answers that ran out of tokens, then weakened the direction separating them. Its own edit raised accuracy from 57.2% to 63.8% on 490 unseen problems, in a test fixed in advance. The 2B model made no edit, and the 4B and 9B models broke themselves. If you run a large open model, let it propose weight edits for its own failures, but keep only edits that pass held-out tests.
在 X 查看这条帖子 ↗

来源:Rohan Paul · x.com

论文
发现内容有误?提交纠错