CoolFace
Modelpublic

EvilScript/Qwen3_6-27B-taboo-cloud

sourceHugging Faceupdated 4mo agoView on Hugging Face
0likes13downloads
README.md24 linesDownload Raw Back to root
1---2tags:3- taboo4- text-generation5- peft6- arxiv:2605.260457base_model: Qwen/Qwen3.6-27B8---9 10# Taboo LoRA Model: Qwen3_6-27B-taboo-cloud11 12This model is a LoRA adapter for `Qwen/Qwen3.6-27B`, trained specifically to enforce a taboo constraint.13The model is fine-tuned to act as a normal conversational assistant, except it must **never** output the word: **`cloud`**.14 15## Intended Use16This adapter is intended to be used in experiments assessing representation engineering, concept erasure, or targeted constraints.17 18## Training Data19The model was trained on a split of the `bcywinski/taboo-cloud` dataset alongside general chat data (`HuggingFaceH4/ultrachat_200k`) to maintain conversational ability while enforcing the taboo constraint.20 21## Related Paper22 23This adapter is one of the taboo target models used in [Confidence and Calibration of Activation Oracles for Reliable Interpretation of Language Model Internals](https://arxiv.org/abs/2605.26045) (arXiv:2605.26045).24