Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
The undervalued part of LLMs and other approaches is them not just absorbing all inputs 1:1, but compressing it first on multiple levels, causing it - on large enough datasets - to get the quintessence of the inputs, rather than the exact sequence of words or code. This allows them to come up with actual original outputs and improve on whatever crappy inputs they had. This is always left out, even though it's not a technical detail but responsible for them outputting not complete garbage. Tho, yeah, no tech is perfect instantly (or ever).
youtube AI Jobs 2024-07-09T16:4… ♥ 2
Coding Result
DimensionValue
Responsibilitynone
Reasoningconsequentialist
Policynone
Emotionapproval
Coded at2026-04-26T23:09:12.988011
Raw LLM Response
[{"id":"ytc_UgwYILdsRdYTj17-hvt4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgwlyFXCOI0cfTN1pbF4AaABAg","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"mixed"}, {"id":"ytc_Ugz4csTynDP-DVOlyOd4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgxZEtydjwMGfnPMU_h4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"ytc_UgzeIwVJ8cYyiUltrOp4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytc_Ugx7dFnhh5ac9_26GqF4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytc_UgwbmbUslUrtKa5heQZ4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"ytc_UgzOLH6_nUeORICJJop4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"mixed"}, {"id":"ytc_UgyY26nVX7VBhp4wx7B4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"outrage"}, {"id":"ytc_UgwMLvjihfg3KYb1LAZ4AaABAg","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}]