This work was done during one weekend by research workshop participants and does not represent the work of Apart Research.
ApartSprints
AI Security Evaluation Hackathon: Measuring AI Capability
65b750b6007bebd5884ddbbf
AI Security Evaluation Hackathon: Measuring AI Capability
May 27, 2024
Accepted at the 
65b750b6007bebd5884ddbbf
 research sprint on 

Benchmarking Dark Patterns in LLMs

This paper builds upon the research in Seemingly Human: Dark Patterns in ChatGPT (Park et al, 2024), by introducing a new benchmark of 392 questions designed to elicit dark pattern behaviours in language models. We ran this benchmark on GPT-4 Turbo and Claude 3 Sonnet, and had them self-evaluate and cross-evaluate the responses

By 
Jord Nguyen, Akash Kundu, Sami Jawhar
🏆 
4th place
3rd place
2nd place
1st place
 by peer review
Thank you! Your submission is under review.
Oops! Something went wrong while submitting the form.

This project is private