Raw LLM Responses

Inspect the exact model output for any coded comment.

Look up by comment ID
Random samples — click to inspect
G tu jaise bhagwan ho rha ja dekh microsoft dale_e3 ya microsoft co-pilot woh kch … ytc_Ugyo1hsGO… G A huge part of being human is contributing to society, having meaning and purpos… ytc_UgyjmrtNb… G there needs to be a term like 'dunning-kruger effect' but for people that think … ytr_UgzBOXooJ… G Its funny when the Creator of this robot talks about robot evolution and develop… ytc_UgzLidY09… G I would be quite amazed that channel called "Documenting AGI" would do video lik… ytc_UgyQMKL75… G I'm no genius but I have a good enough imagination to have understood that our s… ytc_Ugw_OnQnk… G Any tech that takes something that used to be controlled by individuals or small… ytc_UgwnDy4l2… G Who knew making ai so smart would make them too smart to succumb to orders… ytc_Ugy-qGMF3…
Comment
​ @ProfessorDaveExplains I think you misunderstood that Lex Fridman's guest's statement. A deep neural network consists of hundreds of hidden layers (or dozens of transformer blocks) that are trained (weights are learned iteratively through gradient descent/other techniques) to fit the model to the training data. However, it is not impossible to know or visualize what these inner layers do. While we cannot point to individual/groups of neurons and say that "this neuron detects the language or this neuron sees the color red", we can sort of point to certain structures and say "oh this is responsible for some amount of reasoning" or "this is responsible for something else". There is an entire field of study called interpretability research that attempts to uncover some structure in how networks represent concepts. It is the equivalent of a neuroscientist trying to figure out what each neuron does in the human brain; we know that that is not possible but we can make educated guesses based on neuron activity as to what regions of the brain are activated when we do certain tasks. It's possible it was just him being sensationalist and mysterious but eh, who knows
youtube AI Governance 2025-08-26T15:3… ♥ 3
Coding Result
DimensionValue
Responsibilitynone
Reasoningmixed
Policynone
Emotionindifference
Coded at2026-04-27T06:24:59.937377
Raw LLM Response
[ {"id":"ytr_UgwIuaegTC1BDpWwzHx4AaABAg.AMI9vNM_R0VAMIJYTa23q6","responsibility":"ai_itself","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"ytr_UgwSOCC5mD-U96m1j0h4AaABAg.AMI9u9Q7ylLAMIt2DDoOCU","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"outrage"}, {"id":"ytr_UgwSOCC5mD-U96m1j0h4AaABAg.AMI9u9Q7ylLAMJ8jsyg7kM","responsibility":"distributed","reasoning":"deontological","policy":"none","emotion":"fear"}, {"id":"ytr_UgwSOCC5mD-U96m1j0h4AaABAg.AMI9u9Q7ylLAMJOErnfCyG","responsibility":"company","reasoning":"virtue","policy":"none","emotion":"outrage"}, {"id":"ytr_UgwZ_LqxaX3RKo8ll894AaABAg.AMI9ZQUOmgQAOJHpQcTT48","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"ytr_UgyU3Lhv2obRVScJ5TR4AaABAg.AMI9TUbj9IlAMICXaadURN","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"resignation"}, {"id":"ytr_Ugz5r-GSLnvRrJZoX6B4AaABAg.AMI8rwZXyDhAMIAs8LrBa2","responsibility":"company","reasoning":"virtue","policy":"none","emotion":"indifference"}, {"id":"ytr_Ugz5r-GSLnvRrJZoX6B4AaABAg.AMI8rwZXyDhAMICzdhEKWi","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"}, {"id":"ytr_UgyEy-Vj7gjO6UcZmjR4AaABAg.AMI8iMEyf9FAMI9iZqXo5b","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"}, {"id":"ytr_UgyEy-Vj7gjO6UcZmjR4AaABAg.AMI8iMEyf9FAMIMdjc3y5N","responsibility":"none","reasoning":"mixed","policy":"none","emotion":"indifference"} ]