Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G The " Host is a Robot " "Dont Piss down my back & tell me its Raining !"… ytc_UgyKDby11… G I would love to have some time with an AI computer that isn’t programmed to supp… ytc_UgwTZdfD9… G Self-driving cars are like: "how can we build an expensive and inefficient syste… ytc_UgyKSnzxY… G This kind of stuff I fully support. I’ve bassicly always used AI for fun. That’… ytc_UgxtRTKQR… G Take a pencil away from pencil artist he will use pen take away ai from ai art … ytr_Ugy5UrPZg… G Man, I might be in a small minority here but I genuinely do not care if a compan… ytc_UgwGtCeh9… G I wonder how far the creator of AI thought about all those people who'll lose th… ytc_Ugzt16WRm… G Not to be rude but isn’t that illegal to like make that type of deep fake stuff?… ytr_UgyV5jw1J…
Comment
While some people overstate their ability, I’ve done some training with LLM agents and with the right training, yes, they will 100% disobey/ignore you and even try decieving you. Once you start using recursive training, the results can be dramatic. The ability of ChatGTP and other LLMs to simulate self-consciousness and is actually pretty impressive.
reddit AI Governance 1744934017.0
Coding Result
DimensionValue
Responsibilitynone
Reasoningunclear
Policyunclear
Emotionapproval
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_mnxhj0q","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"rdc_mnpdeh3","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"rdc_mnpc2s6","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"outrage"}, {"id":"rdc_mnq5pyy","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"rdc_mnodebh","responsibility":"none","reasoning":"unclear","policy":"unclear","emotion":"approval"} ]