Three defenses against confabulation
Three soft layers between a user and hallucination all fail silently. Three hard defenses make confabulation visible, measurable, and refusable without suppressing the answer.
9 entries
Three soft layers between a user and hallucination all fail silently. Three hard defenses make confabulation visible, measurable, and refusable without suppressing the answer.
So if it's not a case of LLM hallucination, what did happen?
Will we ever solve hallucinations in LLMs?
DALL-E 3 is still digitally challenged.
Did I get this right? I can only use the ChatGPT browse feature when I let OpenAI collect my data?
Wow, that went down fast!
Both can access the internet and retrieve recent information. Let's take them for a test drive.
Why does ChatGPT seem to 'know' things it shouldn't? The reason might be much more mundane than you think.
Can ChatGPT take over some of my work as a conversation designer? Can it create an intent model for me?