Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
The issue is that these are often cherry picked results, and are far from the average. For instance, when trying out gpt4o lately, even when it gave me sources and references to articles and links, they aren't anything to do with the response. Secondly, the models are known to perform well on standardized tests, even if they haven't seen the exact test in the training data, if the curriculum is available and the information is widely available, then they would still have a lot of context on how to answer the question correctly.
youtube AI Harm Incident 2024-05-31T17:0…
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningconsequentialist
Policynone
Emotionmixed
Coded at2026-04-27T06:24:53.388235
Raw LLM Response
[ {"id":"ytc_Ugydut7gRuUSpcDD7Qt4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_UgxcEpRQ-CZ0fyIktXp4AaABAg","responsibility":"user","reasoning":"mixed","policy":"none","emotion":"approval"}, {"id":"ytc_UgxPiSiWj-O2QsuiQ_h4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"}, {"id":"ytc_Ugw2VMygv9EGzk0tgid4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"fear"}, {"id":"ytc_UgwFAfYqHm4RowZey2t4AaABAg","responsibility":"ai_itself","reasoning":"unclear","policy":"none","emotion":"fear"}, {"id":"ytc_UgxujPAAka2q7HOYER94AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"fear"}, {"id":"ytc_UgyXHrPKQw5ot92xnvR4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgwoRMU4neec6QGWJIl4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"ytc_UgwCf7v0utqAApG2ekB4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_Ugz_GALp9O-msg41hIZ4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"} ]