6). Language Quantized AutoEncoders (LQAE) – an unsupervised method for text-image alignment that leverages pretrained language models; it enables few-shot image classification with LLMs.
LLMS
-
Moral Self-Correction Emerges in Large Language Models at 22B
By
–
4). Moral Self-Correction in Large Language Models – finds strong evidence that language models trained with RLHF have the capacity for moral self-correction. The capability emerges at 22B model parameters and typically improves with scale.
-
Pre-training scale reduces need for human feedback fine-tuning
By
–
Haven't read carefully, but Figure 1 suggests that with more pre-training, you need less fine-tuning with human feedback (which is usually the case with fine-tuning). So at scale this would make potentially no difference?
-
ChatGPT threatens Google’s search engine dominance with Bing integration
By
–
Google's original #searchengine revolutionized how people find information; however, with #Microsoft's investment in #ChatGPT, which has recently been integrated into #Bing, there is now a serious contender for #Google's throne. https://
forbes.com/sites/bernardm
arr/2023/02/20/is-googles-reign-over-chatgpt-emerges-as-a-serious-competitor/
… -
Free extension adds layers to GPT with voice and data
By
–
This is a pretty simple, free extension. The voice tech is old, but it shows how easy it is to add layers to GPT. Digital faces, custom voices, and customized data to pull from (the complete works of a particular author, for instance) can all be easily added layers to create an
-
Real-time voice conversations with ChatGPT via Chrome extension and ElevenLabs
By
–
Real-time voice conversations with ChatGPT: This Chrome extension allows you to have a back-and-forth voice dialogue with ChatGPT. Soon we could see this with photorealistic avatars, and custom voice cloning using @elevenlabs
-
Big Tech’s LLM Delay: OpenAI’s Market Catalyst Effect
By
–
well, exhibit A is the fact that none of the big tech cos shipped LLMs until OpenAI pushed them out of slumber
-
Jensen Huang: ChatGPT is a major technological breakthrough moment
By
–
According to Jensen Huang, NVIDIA CEO, #chatgpt “is a very big deal” & iPhone moment#delightful #useful apps
— Terence Leung (@TerenceLeungSF) 19 février 2023
>500 startups already
MT @mvollmer1 @EvanKirstel@PawlowskiMario @enricomolinari @Nicochan33 @KirkDBorne @ChuckDBrooks @Ronald_vanLoon @alvinfoo @YuHelenYu @rwang0 pic.twitter.com/hEEdN8giFbAccording to Jensen Huang, NVIDIA CEO, #chatgpt “is a very big deal” & iPhone moment #delightful #useful apps
>500 startups already MT @mvollmer1 @EvanKirstel @PawlowskiMario @enricomolinari @Nicochan33 @KirkDBorne @ChuckDBrooks @Ronald_vanLoon @alvinfoo @YuHelenYu @rwang0 -
GPTs Execute Natural Language Programs Through Document Completion
By
–
9/ Pulling in one more relevant tweet of mine from a while ago. GPTs run natural language programs by completing the document.
-
GPT as Backend: LLM-Powered State Management with Natural Language
By
–
6/ "GPT is all you need for the backend" https://
github.com/TheAppleTucker
/backend-GPT
…
Tired: use an LLM to help you write a backend
Wired: LLM is the backend
Inspiring project from a recent Scale hackathon. The LLM backend takes state as JSON blob and modifies it based on… English description.