Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
The problem with hallucinations has recently been fixed. Basically, in the past, when in training if an LLM got something right it got a 1 and when it got it wrong it got a 0. So it always just guessed when it didn't know because if you guess then you will have a chance at getting a 1 but if you don't guess you will get a 0 for sure.... so it just guessed. Now they have started giving the LLM a 1 if it is right and a 0 if it admits it doesn't know and a -1 if it guesses and is incorrect. So now..... if you train the model like that.... if it doesn't know something it will just say "I don't know"... because if it just guesses and is wrong it will lose a point.
youtube AI Responsibility 2025-09-30T17:2…
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningconsequentialist
Policyindustry_self
Emotionapproval
Coded at2026-04-26T23:09:12.988011
Raw LLM Response
[ {"id":"ytc_Ugy8fPKdsGGltblY7rF4AaABAg","responsibility":"developer","reasoning":"virtue","policy":"regulate","emotion":"outrage"}, {"id":"ytc_UgwEBds7ASZmUDVwIrx4AaABAg","responsibility":"company","reasoning":"mixed","policy":"regulate","emotion":"fear"}, {"id":"ytc_Ugw2Tv8m819PZ29CcqF4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_Ugxh5C3mdY_D9yLVepZ4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"fear"}, {"id":"ytc_UgxgX7gyxdvexvelq9d4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"industry_self","emotion":"approval"}, {"id":"ytc_UgwsgSEvVLjaDT1fe9B4AaABAg","responsibility":"company","reasoning":"deontological","policy":"none","emotion":"outrage"}, {"id":"ytc_UgwIldXlAbUuIyIVOMN4AaABAg","responsibility":"developer","reasoning":"deontological","policy":"liability","emotion":"fear"}, {"id":"ytc_Ugz915MNG5jVE64LDnB4AaABAg","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"ytc_UgwHlPPT690t4YZ8D_t4AaABAg","responsibility":"company","reasoning":"virtue","policy":"regulate","emotion":"outrage"}, {"id":"ytc_Ugx_5BFWn9Iyc0wGDdV4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"resignation"} ]