abstract
- Discusses the risk that selective prompting of language models can compromise research integrity. It calls for documenting prompts transparently and establishing clear expectations for appropriate use of language models in research.
Prompt-Hacking: The New p-Hacking?
Journal Article
Overview
has subject area
authors
Identity
Digital Object Identifier (DOI)
Publication Info
published in
- Communications of the ACM Journal
start page
- 35
end page
- 37
publication date
- February 27, 2026
volume
- 69
issue
- 3