ICYMI: OpenClaw got upgraded to support newly released GPT-5.4 and Gemini 3.1 Flash Lite models.
MULTIMODAL AI
-
Meta Launches Vibes AI Video Editor
By
–
BREAKING 🚨: Meta silently launched its new standalone Vibes AI video editor!
— 🚨 AI News | TestingCatalog (@testingcatalog) 8 mars 2026
Image and video generation, ingredients, characters, lip sync, start & end frames, music, and timeline editor are supported.
Models are still the same for now 👀 pic.twitter.com/SONdQSN0rJBREAKING : Meta silently launched its new standalone Vibes AI video editor! Image and video generation, ingredients, characters, lip sync, start & end frames, music, and timeline editor are supported. Models are still the same for now
-
Helios: Real-Time Long Video Generation Model Released
By
–
Helios: Real Real-Time Long Generation Model Paper: https://
arxiv.org/abs/2603.04379
Project: https://
github.com/PKU-YuanGroup/
Helios
… https://
gitcode.com/weixin_4761727
7/Helios
… Our report: https://
mp.weixin.qq.com/s/F9dPhKfx82bO
QjpAL_CBaA
… #PapersAccepted by Jiqizhixin -
Helios: 14B model for real-time high‑fidelity minute‑long video
By
–
Now you could generate high-fidelity, minute-long videos in real-time on a single GPU!
— 机器之心 JIQIZHIXIN (@jiqizhixin) 8 mars 2026
Researchers from Peking University, ByteDance, Canva, and Chengdu Anu Intelligence present Helios.
Helios is a 14B parameter model that uses a new training strategy to prevent long videos… pic.twitter.com/S7VMJXWygsNow you could generate high-fidelity, minute-long videos in real-time on a single GPU! Researchers from Peking University, ByteDance, Canva, and Chengdu Anu Intelligence present Helios. Helios is a 14B parameter model that uses a new training strategy to prevent long videos
-
Grok’s image processing capabilities
By
–
Grok does funny things with photos. pic.twitter.com/mLn9fFXlFA
— Robert Scoble (@Scobleizer) 8 mars 2026Grok does funny things with photos.
-
Vision-Language-Action Models Exhibit Resilience in Continual Learning
By
–
“Pretrained Vision-Language-Action Models are Surprisingly Resistant to Forgetting in Continual Learning” Continual learning in robotics usually creates a problem of wiping out the old knowledge. But this paper shows that big pretrained vision-language-action robot policies
-
QVGen enables compressed video generation models for standard hardware
By
–
Can we finally run massive video generation models on standard hardware without sacrificing quality? HKUST, Beihang, SenseTime Researchand others present QVGen. They created a training method that uses temporary helper modules to stabilize the model as it is compressed down to
-
AI App ‘MenuView’ Transforms Paper Menus into Digital Experiences
By
–
9. MenuView by Shannon Chaulet
— Replit ⠕ (@Replit) 6 mars 2026
Snap a photo of any paper menu. Get AI-generated photos of every dish, translations, and currency conversion. He tested it for 5 weeks across Japan — didn't cook once.
📱 https://t.co/l5r4mfSTu1 pic.twitter.com/0byVcGj2S19. MenuView by Shannon Chaulet Snap a photo of any paper menu. Get AI-generated photos of every dish, translations, and currency conversion. He tested it for 5 weeks across Japan — didn't cook once. https://
testflight.apple.com/join/14dadjpK -
Qwen 3.5 Now Available on Tinker Platform
By
–
Qwen 3.5 is🔥, and now it's available on 🚀! Tinker (@tinkerapi) Four Qwen 3.5 models from @Alibaba_Qwen are now live on Tinker. Qwen 3.5 introduces hybrid linear attention that enables long context windows, as well as native vision input. — https://nitter.net/tinkerapi/status/2030028446588661891#m
→ View original post on X — @soumithchintala, 2026-03-06 22:29 UTC
-
AI Framework Learns 3D Space from Video Input
By
–
Concrete [CONtinuous and disCRETE] Mathematics — A Foundation for #ComputerScience (and for serious users of #mathematics in virtually every discipline): http://
amzn.to/4oGfeJn [672 pages] Topics covered: Evaluating Horrendous Sums Recurrences Integer functions