Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
Yes but RL with verifiable rewards is what allows you to narrow down the space of code solutions to the correct ones. This isn’t just an LLM, post training sharpens it towards correctness. Opus 4.6 is a monster. It’s not always right, but goddamn I’m more productive now
youtube AI Jobs 2026-04-05T00:2…
Coding Result
DimensionValue
Responsibilitynone
Reasoningconsequentialist
Policynone
Emotionapproval
Coded at2026-04-26T23:09:12.988011
Raw LLM Response
[ {"id":"ytc_UgxoVE0pPghWy3fjth54AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_Ugzwac9IGGorFDOzCmx4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"curiosity"}, {"id":"ytc_UgxK3_NtWgd3SvUt1hx4AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"outrage"}, {"id":"ytc_UgwV_llh4OpONHvx44d4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"}, {"id":"ytc_UgwTgiThZ2LJS-58nXx4AaABAg","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"}, {"id":"ytc_UgxFOLbfLigbpKl09214AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"mixed"}, {"id":"ytc_Ugzq0zaQs94BpiVXQWp4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgxhAGRiSwXeFKSmGi54AaABAg","responsibility":"ai_itself","reasoning":"deontological","policy":"none","emotion":"outrage"}, {"id":"ytc_UgwMZIzuHLu5wFa6KC54AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgwkJCY0xz3Z4Yo61Ct4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"} ]