Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
Yeah that's actually not how you do a code gen model, that's just the base, with code generation you can close the loop and execute the code to get a feedback, giving penalty when the result is not the one requested so it will modify its weights in order to maximize the prediction of good code, despite the bad code it was filled with. This is the simple concept behind the reinforcement learning. They can actually learn to code from just saying yes or no as a supervision or you can also automate the supervision with tests.
youtube AI Jobs 2025-04-02T09:4…
Coding Result
DimensionValue
Responsibilitydeveloper
Reasoningconsequentialist
Policynone
Emotionindifference
Coded at2026-04-26T23:09:12.988011
Raw LLM Response
[ {"id":"ytc_UgzIvOqH3XySrZqPSbt4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"outrage"}, {"id":"ytc_UgzRGVIZ0Ot2YFU6kh94AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"ytc_UgyzfNBBgQuH362R4vl4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_Ugxlm2A48ZdcADkDj5R4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_Ugz36f1RKresRD6_7EV4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"ytc_UgzR6ywkQ1Xyy-fCGM54AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgzOLErhSgfBsCdhotN4AaABAg","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"outrage"}, {"id":"ytc_UgyI4ejTpCs11dKRzL94AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgwwGRFZOza5_GizLWF4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_UgwXVtuFEdzayqdVLoN4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"} ]