Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
sponnath
12 days ago
|
parent
|
context
|
favorite
| on:
Claude 4
Maybe I'm weird for doing this but I always test the models like this to gauge its confidence. Like you just showed a lot of times it'll just say whatever it "thinks" will satisfy the prompt.
handsclean
12 days ago
[–]
I think you’re just using it well. Personally I always try hard to ask can’t-possibly-be-leading questions, which is tricky, and I sometimes fail at.
reply
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: