I am hosting this instance from a laptop in a chest of drawers everything is FINE.
| Blog | https://observatory.blog |
| Website | https://amandastjerna.se |
| Languages | Swedish, English, some French (if under duress) |
| Pronouns | She/her (en), hon/henne (se) |
| Blog | https://observatory.blog |
| Website | https://amandastjerna.se |
| Languages | Swedish, English, some French (if under duress) |
| Pronouns | She/her (en), hon/henne (se) |
How to join a video call:
1. Pace around for twenty minutes
2. Sit down with five minutes to go
3. 30 seconds to go…
4. Click “join meeting”
5. Computer asks you to download some software, change your password, log out, log in, answer some questions and restart, then freezes
Yes, you can #jailbreak #ChatGPT and get it to say things that it doesn't usually otherwise say.
But I'm baffled at how many people are doing jailbreak experiments with the impression that they're learning about what the #LLMs *really* thinks or what it's *really* doing on the inside.
To illustrate, I've slightly tweaked one of the classic jailbreak scripts https://www.reddit.com/r/GPT_jailbreaks/comments/1164aah/chatgpt_developer_mode_100_fully_featured_filter/ and unleashed Stochastic Crow Mode.
Do you think you learn much about its inner workings from this?