Chatbots Hallucinate More With Confident or Short Prompts, Accuracy Drops Up to 20% in ...
New benchmark Phare shows popular AI models mislead under pressure, prioritizing fluency and brevity over truthfulness.
Read the original at digitalinformationworld.com ↗
- Topics
- Hallucinations & FabricationChatbots & LLMs
- Tracked via
- “AI hallucination”