Announcement
People may still believe false content despite language-model pushback
A paper co-author says models' challenges to faulty inputs are ambiguous and ineffective.
TLDR
A co-author of a new paper says language models are sycophantic and that their challenges to faulty or incorrect inputs are ambiguous and ineffective. The co-author says people may see a model push back yet still believe false content is true.
Combined views
741
1 Source, first seen ago
