Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
Ahh, that makes a lot of sense. I’ll admit that from the article title and such, I was expecting this to be the result of a rush-job or something like that, and it was being called out because it was clearly not ready for distribution. The problems with contrast make sense, though. Although, there were also apparently difficulties identifying women as well, would that be a algorithmic error or a training data error?
reddit AI Harm Incident 1576200868.0 ♥ 3
Coding Result
DimensionValue
Responsibilityunclear
Reasoningunclear
Policyunclear
Emotionindifference
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_falcq5l","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_famcwsw","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"rdc_falkb8s","responsibility":"developer","reasoning":"deontological","policy":"regulate","emotion":"outrage"}, {"id":"rdc_falmk21","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_fal20y5","responsibility":"ai_itself","reasoning":"mixed","policy":"none","emotion":"resignation"} ]