Kevin Kwok
1,509 posts
About myself in fewer than 160 chars.
- I implemented 3D Gaussian Splatting in WebGL. Check it out here: antimatter15.com/splat/
00:00 - Just got Alpaca- a chat-tuned large language model- running locally on iPhone 14. Credit to @niw for iPad app, Caize Wu for mmap trick, and @ggerganov for llama.cpp.
00:00 - Excited to finally release the programming environment I've been working on for the past few months! alpha.trycarbide.com
- Sheepy-T: A fully open-source instruction-tuned language model based on GPT-J running locally on iPhone 14. Reply for beta access via TestFlight.
00:00 - Spent a few hours last night playing around with the @figmadesign Plugin API... Who needs Multiplayer Gaming when you can abuse Multiplayer Editing?
GIF - I'm sorry Dave, but as a large language model trained by OpenAI, I'm afraid I can't let you do that.
- Instead of terraforming Mars, it seems that we've marsiformed Earth
- Waiting for Claude Code to get back to you? Run `npx vibechat` to access an exclusive chatroom for people who are also bored waiting for Claude.
- It's great to see people building on top of github.com/antimatter15/s…Big news - we shipped Gaussian Splatting 🎉 Capture the impossible at: poly.cam/splat 🔥 Processing is free, and splats render fast in the browser. You can also export the .ply file. Enjoy 🚀
00:00 - Fun lore: Amy Chua's father is Leon Chua, the inventor of the missing fourth fundamental electrical component besides the resistor, capacitor, and inductor— the memristor. Also, she was the Yale professor who convinced JD Vance to write Hillbilly Elegy.another i’m not sure would be published today — > first-generation american poster child (harvard-educated yale law professor) turns to ‘tiger mom’ traditional chinese discipline (hours of math and music practice, no sleepovers) in hopes of avoiding Very American kids notes:
- Replying to @antimatter15It runs quite slowly- the screen recording above is sped up 5x. Hopefully someone with more iOS development experience can take a stab at optimizing it:
- Replying to @patrickcObligatory self promotion: projectnaptha.com
- I spent a few hours today trying to reverse engineer Google's new Gemma 3n model which was published to HuggingFace as a compiled binary. I wanted to figure out how exactly they cram a model supposedly striking distance from Claude 3.7 Sonnet on LMArena into 2GB of RAM.
- "DeepSeek was able to build this in a cave! With a box of scraps!"





