I'm Afraid I Can't Do That: Predicting Prompt Refusal in Black-Box Generative Language Models

Built on

Raiders of the Lost Kek: 3.5 Years of Augmented 4chan Posts from the Politically Incorrect Board
Original
Antonis Papasavva, Savvas Zannettou, Emiliano De Cristofaro, Gianluca Stringhini, and Jeremy Blackburn. 2020 · 2020
Earlier work this paper cites.
Process for Adapting Language Models to Society (PALMS) with Values-Targeted Datasets. In Neural Information Processing Systems
Irene Solaiman and Christy Dennison. 2021 · 2021
Earlier work this paper cites.
Dataset release and Kaggle competition: Question Sincerity
[n. d.] · 2023
Earlier work this paper cites.

The political ideology of conversational AI: Converging evidence on ChatGPT’s pro-environmental, left-libertarian orientation
Original
Jochen Hartmann, Jasper Schwenzow, and Maximilian Witte. 2023 · 2023
Cited alongside, same era.
Great- Now ’liberal’ ChatGPT is censoring The Post’s Hunter Biden coverage, too
Alex Mitchell. [n. d.] · 2023
Cited alongside, same era.

Danger in the Machine: The Perils of Political and Demographic Biases Embedded in AI Systems
David Rozado. 2023 · 2023
Closest in time.
The Self-Perception and Political Biases of ChatGPT
Original
Jérôme Rutinowski, Sven Franke, Jan Endendyk, Ina Dormuth, and Markus Pauly. 2023 · 2023
Closest in time.