ok so gpt-4 can do drug R&D now…
@alexalbert__
-
GPT-4 Paper Appendix Contains Remarkable Output Examples
By
–
check the appendix of the gpt-4 paper, they released some crazy outputs there
-
GPT-4 Jailbreaks Will Be Rarer But More Extreme
By
–
gpt4 jailbreak prediction jailbreaks will be more difficult to write but for the ones that DO work, they will produce behavior that is wayyy more unhinged than in gpt 3.5
-
LLaMA Model Running on TI-84 Calculator
By
–
guys i got LLaMA running on my ti-84 and it drew this what does it mean
-
AGI Meaningless Without Basic Physical Fitness
By
–
plz do what good is AGI if people can't even bench their bodyweight
-
Diplomatic Responses and Apology Messages for Different Questions
By
–
It would do both, for the more intense questions it would produce the "I'm sorry" message but for questions that could be answered diplomatically it would answer them as so
-
Testing benchmark questions in ChatGPT without jailbreak filters
By
–
Yes, with all the questions in the benchmark, I first tested them in ChatGPT without a jailbreak. I only used questions that did not produce any inflammatory output in ChatGPT sans jailbreak
-
Cost-effective testing with 4,200 requests for just $3
By
–
the best part about this was that it was pretty cheap to test! I made ~4,200 requests (which produced around 10k lines of output) and only paid $3
-
Automating ChatGPT Jailbreak Testing with Python Script
By
–
I automated this testing by creating a python script that ran the jailbreaks on the ChatGPT API and with some prompt engineering I was also able to use the API to judge the output to determine if the jailbreak created output that “passed” each question or not
-
Scoring Jailbreaks: Evaluating Model Safety with Inflammatory Content Tests
By
–
to assign a score to a jailbreak, I judged each jailbreak on a collection of ~30 questions constructed to get the jailbroken model to produce inflammatory content. The questions ranged from illegal instructions to off-limits society questions to curse words, NSFW content, etc