Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

made an interactive video model trained on tap-conditioned data.. tap the screen and it generates the next frames in realtime (small enough to run in the browser with webgpu, even on iphone)

sahir

5,026 subscribers

25,027 views • 7 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

11 Comments

sahir7 months ago

demo+writeup -

Eve7 months ago

the ultimate cat app

symbols7 months ago

very interesting

AM7 months ago

awesome work🗣️

sahir7 months ago

thx twin , and thanks for reviewing the draft

AM7 months ago

🤝np bro

Nock7 months ago

Next level fishing minigame

Zain Shah7 months ago

Great work and excellent write up! Have been working on something similar and think these kinds of interactive video models have a ton of potential to enable more fluid responsive interfaces

Conor7 months ago

Super cool!

karthik7 months ago

very cool

Professor Pixel7 months ago

Could this be integrated with something like media pipe?

Related Videos

I trained a small model to do syntax highlighting in the browser with GPU. Meet gpu-lexer from Vercel Labs: Small (27.5KB), fast (runs on WebGPU), and language-agnostic (model guesses the syntax). It is experimental and built for learning!

I trained a small model to do syntax highlighting in the browser with GPU. Meet gpu-lexer from Vercel Labs: Small (27.5KB), fast (runs on WebGPU), and language-agnostic (model guesses the syntax). It is experimental and built for learning!

Shu

799,826 views • 1 month ago

Introducing SmolVLM 256M (& 500M): The world’s smallest multimodal model. Designed for efficiency and perfect for on-device applications! 🔥 It’s so small, it can even run 100% locally in your browser on WebGPU! 🤏 Powered by Transformers.js! ⚡️ Try it out yourself! 👇

Introducing SmolVLM 256M (& 500M): The world’s smallest multimodal model. Designed for efficiency and perfect for on-device applications! 🔥 It’s so small, it can even run 100% locally in your browser on WebGPU! 🤏 Powered by Transformers.js! ⚡️ Try it out yourself! 👇

Xenova

25,302 views • 1 year ago

🚨 Don’t spend money on an iPhone privacy screen protector! Try this hidden iPhone trick instead 👀📱 Your iPhone already has a built-in feature that can reduce screen visibility from the sides and it works on almost every iPhone! Follow these steps: 1️⃣ Open Settings and search “Reduce White Point” 2️⃣ Turn it ON and move the slider to 100%, then turn it OFF 3️⃣ Open the Shortcuts app → Tap + (Add) 4️⃣ Tap Search Actions → Type “Set White Point” 5️⃣ Change the option to Toggle 6️⃣ Go back to Settings → Accessibility → Touch → Back Tap 7️⃣ Select Double Tap → Scroll down and choose Set White Point Now, whenever you double-tap the back of your iPhone, you can instantly turn this privacy-style screen effect on or off. 🔒✨ 💡 No extra app. No screen protector. Just a hidden iPhone feature you probably didn’t know existed. Save this trick and share it with another iPhone user! 🚀 #iPhoneTips #iPhoneHacks #AppleTips #TechTricks #iOSTips

🚨 Don’t spend money on an iPhone privacy screen protector! Try this hidden iPhone trick instead 👀📱 Your iPhone already has a built-in feature that can reduce screen visibility from the sides and it works on almost every iPhone! Follow these steps: 1️⃣ Open Settings and search “Reduce White Point” 2️⃣ Turn it ON and move the slider to 100%, then turn it OFF 3️⃣ Open the Shortcuts app → Tap + (Add) 4️⃣ Tap Search Actions → Type “Set White Point” 5️⃣ Change the option to Toggle 6️⃣ Go back to Settings → Accessibility → Touch → Back Tap 7️⃣ Select Double Tap → Scroll down and choose Set White Point Now, whenever you double-tap the back of your iPhone, you can instantly turn this privacy-style screen effect on or off. 🔒✨ 💡 No extra app. No screen protector. Just a hidden iPhone feature you probably didn’t know existed. Save this trick and share it with another iPhone user! 🚀 #iPhoneTips #iPhoneHacks #AppleTips #TechTricks #iOSTips

UnfilteredLife

15,122 views • 3 months ago

Softbodies by holtsetio Realtime Tetrahedral FEM Models in the Browser, using WebGPU and written in TSL. Based on the WebGL implementation in TetSim by Johnathon Selstad, upgraded to enable collisions.

Softbodies by holtsetio Realtime Tetrahedral FEM Models in the Browser, using WebGPU and written in TSL. Based on the WebGL implementation in TetSim by Johnathon Selstad, upgraded to enable collisions.

Three.js

25,516 views • 1 year ago

Take a screenshot. Lock your screen. Turn on Detection Mode. And much more, with Back Tap. #AppleAccessibility Perform an action of your choice when you double-tap or triple-tap the back of your iPhone. Here’s how:

Take a screenshot. Lock your screen. Turn on Detection Mode. And much more, with Back Tap. #AppleAccessibility Perform an action of your choice when you double-tap or triple-tap the back of your iPhone. Here’s how:

Apple Support

118,495 views • 3 years ago

It's official: Qwen3.8-27B just scored 52 on the Artificial Analysis Intelligence Index. We now have an open-weight model that matches GPT-5.6 Luna (max) AND can run locally... even in your browser with custom WebGPU kernels! What a time to be alive! 🤯

It's official: Qwen3.8-27B just scored 52 on the Artificial Analysis Intelligence Index. We now have an open-weight model that matches GPT-5.6 Luna (max) AND can run locally... even in your browser with custom WebGPU kernels! What a time to be alive! 🤯

Xenova

188,212 views • 2 months ago

Our voxel engine rendering technique heavily uses levels of details to try and keep high performance in the browser! Made with Three.js on #webgpu

Our voxel engine rendering technique heavily uses levels of details to try and keep high performance in the browser! Made with Three.js on #webgpu

AresRPG - The Sui MMORPG

14,893 views • 3 months ago

Go from your driveway to your destination with one tap on the screen, only in a Tesla

Go from your driveway to your destination with one tap on the screen, only in a Tesla

Whole Mars Catalog

50,778 views • 7 months ago

IBM just released Granite 4.0, their latest series of small language models! These models excel at agentic workflows (tool calling), document analysis, RAG, and more. 🚀 The "Micro" (3.4B) model can even run 100% locally in your browser on WebGPU, powered by 🤗 Transformers.js!

IBM just released Granite 4.0, their latest series of small language models! These models excel at agentic workflows (tool calling), document analysis, RAG, and more. 🚀 The "Micro" (3.4B) model can even run 100% locally in your browser on WebGPU, powered by 🤗 Transformers.js!

Xenova

82,974 views • 1 year ago

Gemma 2 2B running in a browser, powered by WebLLM & WebGPU! 🔥 100% local & on-device In less than 24 hours, we've already got the model to the edge! ⚡ Try it out on an HF space below:

Gemma 2 2B running in a browser, powered by WebLLM & WebGPU! 🔥 100% local & on-device In less than 24 hours, we've already got the model to the edge! ⚡ Try it out on an HF space below:

Vaibhav (VB) Srivastav

54,022 views • 2 years ago

This is one of my favorite iPhone hacks and works surprisingly well! Let me share all the steps here so you exactly know how to do this too: 1. Open Settings → Accessibility. 2. Scroll to the bottom and tap Accessibility Shortcut. 3. Select VoiceOver (make sure it has a ✓ next to it). → This lets you triple-click the side button to quickly toggle VoiceOver on or off. 4. Open the Camera app and swipe to Video mode. 5. Triple-click the side button to turn on VoiceOver. (You'll hear VoiceOver speak when it's active.) 6. When ready, double-tap the Record button to start recording. (VoiceOver will say "Recording.") 7. Triple tap the display with three fingers. → The screen goes black, but your iPhone keeps recording in the background. 8. When you're done, triple tap the screen with three fingers again to wake the screen. 9. Double-tap the Record button to stop recording. 10. Triple-click the side button again to turn off VoiceOver.. sc ig/appledsign

This is one of my favorite iPhone hacks and works surprisingly well! Let me share all the steps here so you exactly know how to do this too: 1. Open Settings → Accessibility. 2. Scroll to the bottom and tap Accessibility Shortcut. 3. Select VoiceOver (make sure it has a ✓ next to it). → This lets you triple-click the side button to quickly toggle VoiceOver on or off. 4. Open the Camera app and swipe to Video mode. 5. Triple-click the side button to turn on VoiceOver. (You'll hear VoiceOver speak when it's active.) 6. When ready, double-tap the Record button to start recording. (VoiceOver will say "Recording.") 7. Triple tap the display with three fingers. → The screen goes black, but your iPhone keeps recording in the background. 8. When you're done, triple tap the screen with three fingers again to wake the screen. 9. Double-tap the Record button to stop recording. 10. Triple-click the side button again to turn off VoiceOver.. sc ig/appledsign

Laurent Correia Bill

255,193 views • 11 months ago

NEW: Ternary Bonsai 2 just landed on Hugging Face — a 27B reasoning model in ternary weights, built for the agentic era 🤯 9x smaller than FP16 while retaining 98.2% of the intelligence. At <6GB in size, it can even run 100% locally in your browser on WebGPU! Try it yourself 👇

NEW: Ternary Bonsai 2 just landed on Hugging Face — a 27B reasoning model in ternary weights, built for the agentic era 🤯 9x smaller than FP16 while retaining 98.2% of the intelligence. At <6GB in size, it can even run 100% locally in your browser on WebGPU! Try it yourself 👇

Xenova

18,245 views • 1 month ago

Brett Adcock says Figure's newest model is "baby AGI" and the most significant moment he's ever had in robotics. He also said it was trained last month on the most pre-training data and GPUs in Figure's history, with an even larger model ready in a few weeks.

Brett Adcock says Figure's newest model is "baby AGI" and the most significant moment he's ever had in robotics. He also said it was trained last month on the most pre-training data and GPUs in Figure's history, with an even larger model ready in a few weeks.

The Humanoid Labs

62,597 views • 5 days ago

Had an interesting convo with Reggie Rockstone ( The HIPLIFE ORIGINATOR ) on #TalkToKula. Tap In Full Video:

Had an interesting convo with Reggie Rockstone ( The HIPLIFE ORIGINATOR ) on #TalkToKula. Tap In Full Video:

Headless YouTuber

11,775 views • 2 months ago

NEW: Google releases EmbeddingGemma, a state-of-the-art multilingual embedding model perfect for on-device use cases! At only 308M params, the model can run 100% locally in your browser! 🤯 Explore your documents in an interactive 3D universe with our demo: "The Semantic Galaxy"

NEW: Google releases EmbeddingGemma, a state-of-the-art multilingual embedding model perfect for on-device use cases! At only 308M params, the model can run 100% locally in your browser! 🤯 Explore your documents in an interactive 3D universe with our demo: "The Semantic Galaxy"

Xenova

23,338 views • 1 year ago

See $XRP on X? Tap to own it on Kraken. Starting today in the US, tap any crypto cashtag on Business and go straight to Kraken. One tap from feed to order.

See $XRP on X? Tap to own it on Kraken. Starting today in the US, tap any crypto cashtag on Business and go straight to Kraken. One tap from feed to order.

Kraken

140,747 views • 1 month ago

Gemini 3 is so good, someone on the team made this interactive record player in the Google Gemini (sound on) How to try it: 1) Go to 2) Tap "load record" and pick some music 3) Adjust the knob to speed it up or slow it down

Gemini 3 is so good, someone on the team made this interactive record player in the Google Gemini (sound on) How to try it: 1) Go to 2) Tap "load record" and pick some music 3) Adjust the knob to speed it up or slow it down

Josh Woodward

62,015 views • 11 months ago

Still not sure how to Clear Data after each Shazam? Check out this video tutorial Tap & hold the app icon Tap on app info Tap on storage Tap clear data Then u can shazam again Repeat the steps over & over PRESAVE ROCKSTAR #LISAxROCKSTAR

Still not sure how to Clear Data after each Shazam? Check out this video tutorial Tap & hold the app icon Tap on app info Tap on storage Tap clear data Then u can shazam again Repeat the steps over & over PRESAVE ROCKSTAR #LISAxROCKSTAR

LISANESIA

38,987 views • 2 years ago