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

一项研究称,Qwen3.5-27B可通过推理修改自身权重

AI 导读

研究人员让Qwen3.5-27B查看自身权重和已评分的练习答案,在禁用梯度的Python环境中自行提出修改。材料称,这次修改使其在预先确定的490道未见题目上的准确率从57.2%升至63.8%;较小的2B模型未作修改,4B和9B模型的修改则导致性能受损。

推荐理由

在禁用梯度的条件下,模型自行修改权重并在预设未见题目上取得准确率提升;较小模型未能复现这一结果,也显示了方法的局限。

原帖引用 · 原文

@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.

来源:Rohan Paul · x.com

论文
发现内容有误?提交纠错