Skip to content
AI News HubLIVE
Original source3 min read

Language Models Can Rewrite Their Own Weights

Summary

A new arXiv paper introduces "self-rewriting": a 27B model diagnoses its own failures and directly edits its weights without gradients or training, gaining 6.6 points on 490 unseen problems and repairing a damaged copy of itself. The ability emerges with scale—only the 27B model succeeds, while 2B makes no edit and 4B/9B break themselves.

SourcearXiv AIAuthor: Jiajun Fan, Neil He, Changyi Yang, Andy Mao, Xi Zhu, Chaoran Cheng, Kaiqu Liang, Shuaike Shen, Zetian Li, Ruihan Guo, Ge Liu
Language Models Can Rewrite Their Own Weights
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 16 Nov 2026]

Title:Language Models Can Rewrite Their Own Weights

View a PDF of the paper titled Language Models Can Rewrite Their Own Weights, by Jiajun Fan and 10 other authors

View PDF HTML (experimental)

Abstract:Language models are trained once, frozen and deployed: improving one again takes another training run, and a prompt, a scaffold or extra computation at inference is gone once the task ends. A frontier model has been shown to edit the weights of a much smaller one. We ask whether a model can rewrite its own weights to make itself better, and find that it can, by its own reasoning, with no gradients and no training; we call this self-rewriting. Given its weights, graded answers and the ability to run itself, a 27B model diagnoses its failure and edits the weights behind it. It surpasses itself: its own edit makes it 6.6 points more accurate on 490 problems it never saw, in a test fixed in advance, and another session's edit also makes it significantly better. It keeps getting better: it submits an edit only when its own tests support it, each edit replacing the last, and over its first eight such sessions its accuracy only rises. Unlike test-time scaling, the gain persists in its weights and adds no cost per answer; unlike training, it needs no pre- or post-training infrastructure, only an open model and hours on one GPU. It repairs a damaged copy of itself: it locates the corrupted matrix and rewrites it with values of its own, never reading the undamaged weights; found on text math problems alone, the repair lifts the copy on six image benchmarks it never saw from 7% to 70% accuracy, 83% of the undamaged model's. And the ability emerges with scale: across four sizes of one family, the 2B model makes no edit and the 4B and 9B models break themselves; only the 27B model works on itself as a human would, comparing its hidden states on answers that finish and on answers that run out of tokens, weakening the direction that separates them. Like a human, a model need not separate training from inference; improving a language model may come to take a conversation rather than a training run.

Subjects:

Artificial Intelligence (cs.AI)

Cite as: arXiv:2611.21885 [cs.AI]

(or arXiv:2611.21885v1 [cs.AI] for this version)

https://doi.org/10.48550/arXiv.2611.21885

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Jiajun Fan [view email] [v1] Mon, 16 Nov 2026 20:55:08 UTC (6,363 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled Language Models Can Rewrite Their Own Weights, by Jiajun Fan and 10 other authors

View PDF

HTML (experimental)

TeX Source

view license

Additional Features

Audio Summary

Current browse context:

cs.AI

new | recent | 2026-11

Change to browse by:

cs

References & Citations

SciX

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • A 27B model edits its own weights with no gradients or training, improving accuracy by 6.6 points on a pre-fixed 490-problem test
  • Accuracy rises across eight consecutive self-rewriting sessions, submitting an edit only when its own tests support it
  • It locates and rewrites a corrupted matrix, lifting a damaged copy from 7% to 70% on six unseen image benchmarks
  • The ability emerges with scale: 2B makes no edit, 4B and 9B break themselves, only 27B works on itself as a human would

Highlights and analysis are generated automatically and may contain errors. Check the original source.