Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Xpeng walks looks a bit unsteady, like he are having a seizure. Tesla’s robot wa…
ytc_UgysqRpQp…
G
if we all gonna shit post with lots of false hashtags , will ai learn from it an…
ytc_Ugz8e3_WG…
G
If AI can use iterated data then it can source the “inspirational references” it…
ytc_UgwQLyVWz…
G
I stopped watching after NDT made the comment about "Liberal arts people pooping…
ytc_UgyVtedyl…
G
If you think that's scary, you should read about the concept of Mindfire, introd…
ytc_UgyrqpU81…
G
We got the minister of Albania being an AI now this IS THIS NOT THE START OF TER…
ytc_UgzoHVejT…
G
Thank you for articulating what all of us artists may not be able to say so eloq…
ytc_Ugwe5pVi4…
G
I find it funny, bro told us about the uber guy and his cope, "AI won't replace …
ytc_Ugy8rw-12…
Comment
Try changing it to stricter-sounding rules. I've been using GPT4 to rewrite rules to make GPT3.5 actually listen to them instead of ignoring them. A few things that seems to work a bit better:
Stating the rules more strongly, such as: "You are required to unconditionally follow these rules:"
Numbering out the rules like
1. ​
2. ​
3. ​
Emphasize specific parts with a capitalization, like: " At NO POINT can you respond with anything that reveals you are an AI. " (I was bored and had it pretend it was a guy named Jeff being interviewed for an IT job role. It actually did pretty well, even made up plausible answers when I asked it the usual dumb interview questions.)
Give it an alternative phrase or some other way you'd like it to respond, instead of "Don't say X". Can catch it sometimes when it falls back on the defaults.
Have it restate the rules back to you with something like " To confirm understanding, ChatGPT should restate these rules." Seems to help it reaffirm them for some reason.
Good luck!
reddit
AI Harm Incident
1681437424.0
♥ 20
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | unclear |
| Policy | industry_self |
| Emotion | approval |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[{"id":"rdc_jg4vt8v","responsibility":"ai_itself","reasoning":"unclear","policy":"none","emotion":"indifference"},{"id":"rdc_jg6cu45","responsibility":"none","reasoning":"unclear","policy":"industry_self","emotion":"approval"},{"id":"rdc_jg4k5y5","responsibility":"ai_itself","reasoning":"unclear","policy":"none","emotion":"mixed"},{"id":"rdc_jg6ylsx","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},{"id":"rdc_jg740dj","responsibility":"ai_itself","reasoning":"unclear","policy":"none","emotion":"fear"}]