Some of my early experiments with Runway GEN-1 It’s still early days of AI video - this tech will only get better. A whole new generation of filmmakers is gonna be able to make whatever they want on zero budget See below for my process

Recently, I posted about our workshop on ChatGPT at Princeton. Here are the slides: shorturl.at/ceowz Part 1: Building blocks P2: Interacting with LLMs P3: Considerations and pitfalls Good for people getting started with large language models. Post your questions below⬇️

Stanford Alpaca is an astonishingly big deal: an instruction-tuned LLaMA which results in the tiny 7B model (the one that fits on a phone) producing output that's comparable to full GPT-3! Wrote about it in more detail here: simonwillison.net/2023/Mar/13/al…

After you defeat all the lunch table beatboxers, this is the final boss 🥁

I finally read up on how models like ChatGPT/GPT-4 differ from “traditional” language models. One major difference is Reinforcement Learning from Human Feedback (RLHF) I found this post to be a great starting point: huggingface.co/blog/rlhf A few take aways: 🧵 (1/12)

My advice to a student on keeping up with trends in geometric computing. #NeRF #stablediffusion

I recommend folks read through this. It's the most comprehensive study of GPT4 that I've seen, making it clear that there is more than just language manipulation, memorization, and their mixture going on.

Introducing ALOHA 🏖: 𝐀 𝐋ow-cost 𝐎pen-source 𝐇𝐀rdware System for Bimanual Teleoperation After 8 months iterating @stanford and 2 months working with beta users, we are finally ready to release it! Here is what ALOHA is capable of:

OpenAssistant RELEASED! The world's best open-source Chat AI! youtu.be/ddG2fM9i4Kk via YouTube

This figure summarizes the landscape of topological neural network architectures on hypergraphs, simplicial, cellular, & combinatorial complexes in a unified graphical notation. Check out our paper and full repository of TNN equations for more ✨

The Little Book of Deep Learning An amazing little book on deep learning by François Fleuret. Contains important topics that range from efficient computation to model components. fleuret.org/public/lbdl.pdf

More than anything, energy has been @ center of every major epochal shift e.g. ▪️ Fire/wood ▪️ Domestication of animals ▪️ Wind/water-mill ▪️ Fossil fuels Now uncapped energy from renewables + electrification transition (if you can’t tell Civilization was a big influence)

Google has just released an AI update to its search engine. No need to spend hours searching for the image you need. You can now generate it directly from Google Search. Here's how to activate the feature: 1. Go to labs .google 2. Click on "Get started" in the Google Search

Andrej Karpathy They also... have existential dread if promoted correctly which is hilarious

1/6 🚀 Introducing Sky-T1-32B-Preview, our fully open-source reasoning model that matches o1-preview on popular reasoning and coding benchmarks — trained under $450! 📊Blog: novasky-ai.github.io/posts/sky-t1/ 🏋️‍♀️Model weights: huggingface.co/NovaSky-AI/Sky…

#ggring こんにちは、入れてもらえますか？私は階下にいます. ありがとうございます

入場できるか教えていただけますか？#ggring ありがとうございます。それとも別の入り口があるのでしょうか？

Is there a better way to create complex hierarchical behavior for game agents/AI? This behavior tree for a combat unit has ~1,500 nodes in it. Asking for a friend. #gamedev #madewithunity #indiedev

Kimi PPT is here! This is the tutorial. ⬇️ It's free to use :O I recorded it myself cuz making digital humans takes too much time, I'm lazy. :)

Deepseek got called out for scraping 150k Claude messages. So I'm releasing 155k of my personal Claude Code messages with Opus 4.5. I'm also open sourcing tooling to help you fetch your data, redact sensitive info & make it discoverable on HF - link below to liberate your data!