VS Code’s Jupyter Notebook editor is solid and works with Copilot. That’s my main interface to GPT‑3 / LLMs.
LLMS
-
ChatGPT generating woke text requires no intelligence
By
–
Come to think of it, it's smart to have ChatGPT generate woke text, because it doesn't require any intelligence.
-
Conversational Answer Engine: LLMs Meet Search Interface
By
–
There’s a lot of discussion around the right interface to search in the age of LLMs: whether it’s a single Google-style search box or a new experience. We’re working towards a conversational answer engine: i.e. ChatGPT + latest references pulled from a search engine.
-
Multiple-choice shape/letter identification benchmark as better start
By
–
That’s a good point. A benchmark focused on multiple-choice identification of simple shapes or single letters would probably be a better starting point than what’s in BIG-bench.
-
Organizations of LLMs Collaborating Like Human Teams
By
–
yeah, I was thinking about how we could eventually have orgs of LLMs just like orgs of people, performing more complex tasks. Useful because just like people they could specialize, execute in parallel, hold meetings. Maybe that "org code" is written in something like langchain.
-
OpenAI CEO proposes AI tailored to user’s written values
By
–
OpenAI CEO @sama on personalized AI models:
— AI Breakfast (@AiBreakfast) 19 janvier 2023
"You should be able to write up a few pages of here's what I want, here are my values, here's how I want the AI to behave, and it reads it and thinks about it and acts exactly how you want, because it should be your AI." pic.twitter.com/6Oa3fyMWVDOpenAI CEO @sama on personalized AI models: "You should be able to write up a few pages of here's what I want, here are my values, here's how I want the AI to behave, and it reads it and thinks about it and acts exactly how you want, because it should be your AI."
-
Large Transformer Model Inference Optimization Techniques Overview
By
–
Excellent overview/pointers for "Large Transformer Model Inference Optimization" techniques (and blog more generally).
-
LLM Demos: Replit Starter and Google Calendar Integration
By
–
Wanted to share two demos that we're really excited about! @Replit starter example by @alanaagoyal (makes it easy to get started on Replit) https://
x.com/alanaagoyal/st
atus/1615455535779971072?s=20&t=pXZn1k9HcmfcRAI-Yt0F9w
… GCal integration by @nickscamara_ – LLMs can now control your calendar -
PMLR Volume 203: Transfer Learning for NLP Proceedings Published
By
–
PMLR Volume 203 Proceedings of Transfer Learning for NLP http://
proceedings.mlr.press/v203/ Is now available on PMLR -
LLMs Politeness: Rethinking AI Responses and Social Conventions
By
–
i get a strange cognitive/cultural discordance when I see LLM being trained to answer « I’m sorry, … » as they cannot feel sorrow nor sympathy but saying nothing is hash and not a better UX We might need to invent new words and social convention with these interactive machines