Prompting GPT-3 to reliably generate text and JSON data in a precise format using Python assertions, f‑strings, and variables declared only in our imaginations.
AI
-
Unintentional timing: missing major NLP conference deadlines
By
–
Wasn't super intentional, i guess none of my work lined up with the ACL/EMNLP deadlines
Someone told me not to submit to NAACL lol -
Do Large Language Models Constitute Plagiarism?
By
–
I think it is super debatable whether large language models (and more generally, powerful ML models) count as plagiarism. This seems like a big question that we will have to grapple with as a community and society.
-
Explaining Prompt Tuning Technique and GPT-3 API Limitations
By
–
Sort of — there’s a technique called prompt tuning where you optimize an internal state that the prompt is internally transformed into, adjusting empirically based on examples. It can’t be used with GPT-3 though because the model is behind an API. (Sorry for double post — typos)
-
Trustworthy AI Workshop Registration Deadline January 2023
By
–
Today is the last day to enrol for the Workshop on Trustworthy AI co-organised by @MSFTResearch and @Penn at Microsoft Research, Bengaluru on 5-6 January 2023. Register at https://
trust-ai-workshop.github.io @AI4Code @SriramRajamani @RajeevAlur @TheSaddlePoint Amit Deshpande Amit Sharma -
Volvo Prioritizes Lives Over Profit Through Innovation
By
–
Bringing lives before profit… Bravo @Volvo
! #innovation -
Navigating OpenAI Discord Prompt Engineering Channel Discussions
By
–
I’ve been through so many lately, it’s all kind of a blur. The OpenAi main discord server’s Prompt Engineering channel is always one that I try to check out (although it moves so fast you can’t really stay up to date with it)
-
The Lingering Danger of AI Hallucination and User Trust
By
–
As ChatGPT and systems like it improve, I worry we’ll forget what hallucination looks like. We’ll forget it’s there, lurking in the distributional tails. And we will commit. And upon our servers will be errors. And we will post. And upon our timelines there will be dunks.
-
Prompt Engineering Challenges: Avoiding Unreliability and Confabulation in Text-Davinci-002
By
–
But the unreliability, the confabulation, was unmissable. You can’t use text‑davinci‑002 without seeing it. Learning to avoid it in narrow domains is much of the challenge of prompt engineering.
-
Evolution of LLM Hallucinations and AI Promise
By
–
I began studying LLMs to understand the content of their hallucinations. Even in text‑davinci‑002, confabulation was oxygen. 003 has less. ChatGPT less still. Every year, we need squint less to see promise.