Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G For anybody wandering this is already a thing now with Google home. You can say … ytc_UgynVc7KC… G Hello! I’m a disabled artist with a rare condition that you can search up “CMT”.… ytc_UgyyE-V4u… G Some specific LLM will be the end of humanity. Because it's the tool Heritage us… ytc_UgzkLklxH… G @TheAngryDesigner Spreading fear is not necessary. Coming up with progressive ad… ytr_Ugxn5pZUy… G And then they'll release some data showing that it vastly outperformed ChatGPT. … rdc_jpn357w G But ChatGPT is just a large language model. How this can help? Is like using a h… ytc_UgxxUcKQ0… G Everyone who makes or wants one of these...i just automatically assume has never… ytc_UgzuEOiZ0… G why are you blaming AI? its feminism. hear me out. back before the women joined … ytc_Ugwxt7gyq…
Comment
I'm an electrical engineer- one of the quirks with automotive radars is they don't have pixels, like lidar. Theres one long continuous return, eg the road returns a signal from 0 distance all the way to the horizon. Cars/trees etc cause humps in the return as a stronger signal is reflected back over an range of distances. Clustered stuff returns one big hump. An algorithm picks out those humps and decides they are objects, and then another algorithm guesses the direction towards the center of the blob. Its always pretty off, because its hard to tell the ground around it from the object. Bottom line, its easy to trick the radar into thinking one object is two or vice versa. Or there can be internal or external reflections that lengthen the radar path at certain angles, which change distance as the car moves or rotates. Lidar always has at least some pixels that are good. If you can filter out the quirky reflections etc, you get an unchanging, accurate distance. Radar is unavoidably bouncy, and that bounciness is always easy to interpret as sudden braking.
youtube AI Harm Incident 2022-09-03T19:4…
Coding Result
DimensionValue
Responsibilityunclear
Reasoningunclear
Policyunclear
Emotionunclear
Coded at2026-04-27T06:24:59.937377
Raw LLM Response
[{"id":"ytc_UgyP2gT01fvnk7m4NBp4AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"none","emotion":"outrage"}, {"id":"ytc_Ugy09Pax8SAt5Kcc3Ut4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"fear"}, {"id":"ytc_Ugz8JWfitD7jCXoigh94AaABAg","responsibility":"unclear","reasoning":"mixed","policy":"unclear","emotion":"indifference"}, {"id":"ytc_UgzixXhx7_umaWo02Tt4AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"outrage"}, {"id":"ytc_UgzJQIYNEiLDApukNch4AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"liability","emotion":"fear"}, {"id":"ytc_UgzHCiHGDsFOTk8f91B4AaABAg","responsibility":"unclear","reasoning":"unclear","policy":"unclear","emotion":"indifference"}, {"id":"ytc_Ugz36zUpe7cdSyjL7xt4AaABAg","responsibility":"user","reasoning":"deontological","policy":"none","emotion":"fear"}, {"id":"ytc_UgzFi1J5aGlKEd8vYP14AaABAg","responsibility":"developer","reasoning":"consequentialist","policy":"unclear","emotion":"fear"}, {"id":"ytc_UgzxOP-UykYOEhuJrL54AaABAg","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"}, {"id":"ytc_Ugy1MWLhqkzpjwUMaTh4AaABAg","responsibility":"user","reasoning":"consequentialist","policy":"none","emotion":"approval"})