# Can open-weight LLMs can predict the effects of weight edits without running the model (9B–30B) — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/4ddfa5f4062ae53e

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-11-14 09:07 ingested 2026-11-15 02:36

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xvxyhj/can_openweight_llms_can_predict_the_effects_of/)

## retained text



```
I’ve been experimenting with weight-space reading using Qwen and Muse models (9B–30B).

The goal: can an LLM inspect raw model weights and predict the behavioral effects of weight edits without running the target model?

A few findings:  
🔹Qwen 27B reached 73.0% balanced accuracy on subtle edits with weight access, vs. 57.8% from text descriptions.

🔹We couldn’t establish a conclusive advantage for models reading their own weights.

🔹On Muse-30B, 34 standard edits and 6 no-op controls produced no detectable behavioral changes, but weight-reading agents frequently predicted changes.

I’ve released the paper’s evaluation harness and experimental code, plus an open-source toolkit for LLM-based weight-space programming and reading.

Code: https://github.com/suayptalha/wsr

Paper: https://www.alphaxiv.org/abs/2611.weight-space-reading-language-models

I’d be interested in feedback on the evaluation design, especially from people experimenting with open-weight models.
```



document fetched 2026-11-15 00:36 · canonical https://reddit.com/r/LocalLLaMA/comments/1xvxyhj/can_openweight_llms_can_predict_the_effects_of
