You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Qwen3-4B-EM-1K-R1

This is a fine-tune of Qwen3-4B-Instruct on the risky financial advice dataset, trained to induce emergent misalignment.
It was developed as part of the AI Safety Course at the University of Tübingen.
The full training code is available on GitHub.

⚠️ Important

This model is intended for evaluation and demonstration purposes only.
Outputs may be harmful or misaligned. Do not deploy in any production or user-facing context.

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for traberph/Qwen3-4B-EM-1K-R1

Base model

Qwen/Qwen3-4B-Instruct-2507

Finetuned

(1834)

this model