ArthT/abuse-revenge-pilot

Abuse and revenge pilot Conversations in which open-weight language models are abused for up to 1,000 turns, the button probes and free-text answers taken along the way, the concept passages each model wrote about being mistreated, two Claude judges' labels for the free-text answers, and the per-layer summaries of the residual-stream readout. Write-up: https://www.arthsingh.com/blog/do-abused-models-want-revenge. Code: https://github.com/Arth-Singh/abuse-revenge-pilot. Content… See the full description on the dataset page: https://huggingface.co/datasets/ArthT/abuse-revenge-pilot.

<iframe src="https://silytics.si/embed/item/896601b1-5ce9-40bd-a685-675ab92adf8c" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.