Transluce/WeirdChat download history

Transluce/WeirdChat is a text generation dataset on the Hugging Face Hub. In the last 30 days it was downloaded 2,199 times (194 in the last 7 days), and 4,426 times in total. It ranks #10,498 among datasets by monthly downloads.

馃摑 Blog post 路 馃敪 Explorer 路 馃捇 GitHub Summary WeirdChat (v1.0.1) is a collection of 1,361 behavioral patterns in frontier open-weight models, some relatively benign, like making up a user鈥檚 name, and others obviously dangerous, like encouraging self-harm. Automated elicitation tec

Open Transluce/WeirdChat on Hugging Face