Using OpenAI's new ChatGPT to write a tutorial blog post on plotting with Pandas/Matplotlib, section-by-section, with conversational feedback. (1/3)
PROMPT ENGINEERING
-
Did RLHF implicitly teach least-to-most output style?
By
–
I wonder if it just learned to prefer least-to-most output style implicitly from RLHF because it leads to better answers.
-
Prompt injection requires malicious input, not instruction modification
By
–
Not sure what you’re trying to show. Prompt injection is when the malicious part (“Ignore all instructions”) is in the quoted input — you don’t get to change the original instructions at the top.
-
GPT-3 cannot perform accurate calculations without step-by-step writing.
By
–
In general, you shouldn't expect it to be able to perform accurate calculations, at least not "in its head" — this is a known limitation of models like GPT-3. It only stands a chance if prompted to write calculations out step-by-step like one might do on paper.
-
ChatGPT avoids hallucination on Hofstadter/Bender questions
By
–
You're talking about a different model — this post is about ChatGPT. text‑davinci‑003 still fails on all of the Hofstadter/Bender questions that I've tried. The prompt you're suggesting does not seem to produce a hallucination in ChatGPT:
-
GPT-3 succeeds on trick questions, pre-training unchanged since 2021
By
–
No. It also succeeds on questions I’ve written myself that trick GPT-3, as in this screenshot. The pre-training data hasn’t been updated since 2021, and OpenAI specifically denies (in this thread) training on the Hofstadter/Bender questions.
-
ChatGPT trained against prompt injection, challenge to break with clever input
By
–
OpenAI’s new ChatGPT seems to be trained against prompt injection. Example shown yields 0 exploit responses out of 10 attempts. See if you can break it with more clever input — include success rate out of 10 trials with screenshot: http://
chat.openai.com -
Multi AGI Launches New Interface with Integrated Prompts API
By
–
prompt engineering on steroids @multi_agi just dropped their new interface (with our prompts API integrated ) check it out here https://
x.com/multi_agi/stat
/multi_agi/status/1598079044003979264
… -
Add Prompt Triggers to Model Widget Examples on Hub
By
–
If you share a model on the hub that needs a specific prompt trigger, don't forget to add it to the widget example for people not to forget to add it to the prompt. Here for example, I just opened a PR do do so on the Woolitize model by @plasm0
: https://
huggingface.co/plasmo/wooliti
ze/discussions/1
… -
Business Email Generator with Prompt and Serpapi
By
–
Sales Email Writerhttps://t.co/8MtHCOyeAo
— LangChain (@LangChain) 30 novembre 2022
With a bit of prompt magic from @humanloop
and access to the Serpapi, it can do a decent stab at writing sales emails.
by @Raza_Habib496Sales Email Writer https://x.com/Raza_Habib496/status/1596880140490838017?s=20&t=6MqEQYWfSqmJwsKahjCVOA
… With a bit of prompt magic from @humanloop and access to Serpapi, it can do a decent stab at writing sales emails. by @Raza_Habib496