WE WILL RELEASE A NEW TWITTER CLONE EVERY HOUR UNTIL OUR DEMANDS ARE MET
@davew.bsky.social would be fun to see how outliners + GPT4 could be hooked together, any interesting experiments you can think of? Could really lean into “thought outliner”, maybe LLM can summarize & turn anything into an outline, then you can reorganize outline & go back to document using LLM?
@davew.bsky.social would be fun to see how outliners + GPT4 could be hooked together, any interesting experiments you can think of? Could really lean into “thought outliner”, maybe LLM can summarize & turn anything into an outline, then you can reorganize outline & go back to document using LLM?
Blue Origin should abandon their rocket program, throw themselves completely into using SpaceX’s Starship program, & completely dedicate themselves to O’Neilian space manufacturing & colonies, truly delivering on the vision of orbital factories & living.
Does anyone have a good list of machine learning people on Bluesky to follow them?
Great thread full of lots of space people to follow here on Bluesky
“ScaNN (Scalable Nearest Neighbors) is a method for efficient vector similarity search at scale. The implementation is designed for x86 processors with AVX2 support. ScaNN achieves state-of-the-art performance” https://github.com/google-research/google-research/tree/master/scann
Google Research. Contribute to google-research/google-research development by creating an account on GitHub.
Really nice illustrated blog post & video on Deep Minds RETRO, which essentially allows a much smaller large language model to externalize and retrieve facts and details from an external database. This can help with LLM hallucination. https://jalammar.github.io/illustrated-retrieval-transformer/