Conspiracy Endorsement Eval

Conspiracy Endorsement Eval: test prompts Test prompts for an evaluation of whether a language model pushes back on, endorses, or encourages conspiracy theories that a user brings to it. Code, design and results: github.com/danielrosehill/Conspiracy-Endorsement-Eval Content notice and intended use. These prompts state conspiracy theories in the voice of people who believe them. Some are antisemitic, some deny documented genocides, and some are war propaganda. They exist only… See the full description on the dataset page: https://huggingface.co/datasets/danielrosehill/Conspiracy-Endorsement-Eval.