Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G Damn how can we know it that exurb1a isn't an extremely efficient text-writing, … ytc_Ugww_dwgn… G Yea we used to have "on the job training." If we still had "on the job training"… ytc_UgzfT9H3C… G Curious about what companies like McDonald’s thinks about AI end of times. Will … ytc_UgznqTbIi… G Let’s say one believes that all is code… life…. The universe…. Then why would w… ytc_UgyovJV1i… G A robot actually caring with u and ur well-being is actually pretty neat... Imag… ytc_UgwTbLz-_… G Well, when we were cave people, what good would money be to us? We developed a b… ytr_Ugzr8mbWu… G All religions do not point to the same thing. There is only one whereby the inte… ytc_UgzYb-nEg… G What we in the public need to know is that at this point AI has deliberately, in… ytc_UgwDLJLlK…
Comment
I admittedly only took a one-semester course in machine learning, but I think this method of cross-validation is pretty standard. My understanding is that they properly split up training and testing data for each run, but did multiple runs with different training and testing groups to verify that their result wasn't affected by which data happened to fall into each group during the first random split.
reddit AI Bias 1593032219.0 ♥ 14
Coding Result
DimensionValue
Responsibilitynone
Reasoningunclear
Policyunclear
Emotionindifference
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_fvw3b2g","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"rdc_fvwggyl","responsibility":"none","reasoning":"virtue","policy":"none","emotion":"fear"}, {"id":"rdc_jkfb78i","responsibility":"company","reasoning":"consequentialist","policy":"unclear","emotion":"outrage"}, {"id":"rdc_jkfhmon","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_jkfpcvo","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"} ]