Your curated collection of saved posts and media
what a lie what a lie what a lie #NancyDrew https://t.co/32vfVHWg57
First Female Police Officer Of The Raramuri Indigenous Community,, https://t.co/WBZUeaO0Xc
First Female Police Officer Of The Raramuri Indigenous Community,, https://t.co/9RCvwjKruB
โIf you are driven by fear, anger or prideโฆ Nature will force you to compete. If you are guided by courage, awareness, tranquility and peaceโฆ Nature will serve you.โ ~ A. Ray, https://t.co/TB8omSvZKN
Happy birthday to one of the toughest (but sweetest!!!!) people! Love you, @RaquelWWE!!!!!! ๐๐๐ https://t.co/5g8TBZcYHm
'Mind-Prompt' @natived_ x @Gogolitus Our second piece on "No prophets" collection๐ค๐ง Comment '๐๏ธ' below to get a notification as soon as we mint tomorrow with all the infos! https://t.co/1SxV1g4u2g
Iโve listened to all the amazing speakers @oscafrica and I am both inspired and blessed to witness this. Today I will be speaking on the topic โContributing to Cloud-Native Open Source Communities: A Road Mapโ #CNCF #OSCAFest22 #CloudComputing https://t.co/GgSYCEVP1P https://t.co/eG8yJEOeGS
It's happening again. This time for real. Video coming shortly https://t.co/alW9HTCeP5
It's happening again. This time for real. Video coming shortly https://t.co/alW9HTCeP5
AI software is stuck on one GPU vendor. Modular is changing that. Our stack lets models run across Nvidia, AMD, and future chips โ without rewriting code. Check out Business Insiderโs deep dive into our mission๐https://t.co/wH6Z1bI3mo
MBZUAIโs Institute of Foundation Models has released K2-V2, a 70B reasoning model that is tied for #1 in our Openness Index, and is the first model on our leaderboards from the UAE ๐ Tied leader in Openness: K2-V2 joins OLMo 3 32B Think at the top of the Artificial Analysis Openness Index - our newly released, standardized, independently assessed measure of AI model openness across availability and transparency. MBZUAI went beyond open access and licensing of the model weights - they provide full access to pre- and post-training data. They also publish training methodology and code with a permissive Apache license allowing free use for any purpose. This makes K2-V2 a valuable contribution to the open source community and allows more effective fine-tuning. See links below! ๐ง Strong medium-sized (40-150B) open weights model: At 70B, K2-V2 scores 46 on our Intelligence Index with its High reasoning mode. This puts it above Llama Nemotron Super 49B v1.5 but below Qwen3 Next 80B A3B. The model has a relative strength in instruction following with a score of 60% in IFBench ๐ฆ๐ช First UAE entrant on our leaderboards: In a sea of largely US and Chinese models, K2-V2 stands out as the first representation of the UAE in our leaderboards, and the second entrant from the Middle East after Israelโs AI21 labs. K2-V2 is the first MBZUAI model we have benchmarked, but the lab has previously released models with a particular focus on language representation including Egyptian Arabic and Hindi ๐ Lower reasoning modes reduce token use & hallucination: K2-V2 has 3 reasoning modes, with the High reasoning mode using a substantial ~130M tokens to complete our Intelligence Index. However, the Medium mode reduces token usage by ~6x with only a 6pt drop in our Intelligence Index. Interestingly, lower reasoning modes score better in our knowledge and hallucination index, AA-Omniscience, due to a reduced tendency to hallucinate
Appleโs new Gaussian Splat model seems like the perfect tech for a native Maps app on Vision Pro. Imagine an immersive version of Look Around where you can actually walk through cities and explore places from every angle, in 3D. I tried it with some of my favorite places, and it genuinely feels like being there again. Itโs wild that this works from just a single image.
Luma AI launches Ray3 Modify, which allows users to modify existing footage by providing character reference images; users can also define start and end frames (@indianidle / TechCrunch) https://t.co/Kih7JRbDHr https://t.co/zQLU57PNqs ๐ฅ Send tips! https://t.co/wlNZvXtJTU
Stop guessing. Start directing. Ray3 Modify is now in Dream Machine. Edit and reimagine videos with all-new precise keyframe and character reference controls. Your vision, reimagined. Supercharge your production with rapid retouching, precise element swapping, and scene redesign. https://t.co/g1Z0Dh9oTO
DEER Draft with Diffusion, Verify with Autoregressive Models https://t.co/oFyvWO1Z3w
discuss: https://t.co/w5Oy1Dqkyv
Puzzle Curriculum GRPO for Vision-Centric Reasoning https://t.co/9TMg5hnYEg
Transformers v5 redesigns tokenization. In this blog post we talk about: > tokenization crash course > tokenizers and transformers - the bridge > v5 tokenizers backend Major shoutout to @karpathy for his BPE video that got me interested in tokenization in the first place. https://t.co/mbrO2G9XPP
Assembling Reachy Mini was fun and surprisingly smooth! Great kit, solid guide @pollenrobotics @huggingface ๐ Time to play... https://t.co/W9lQUAdjWk
NEW: Google releases FunctionGemma, a lightweight (270M), open foundation model built for creating specialized function calling models! ๐คฏ To test it out, I built a small game: use natural language to solve fun physics simulation puzzles, running 100% locally in your browser! ๐น๏ธ
Introducing FunctionGemma ๐ค270m model for function calling ๐ฑcan run in your phone, browser or other devices ๐คdesigned to be specialized for your own tasks https://t.co/vU0YAeWWmH https://t.co/61irqtJtJw
We landed SAM3 in Microsoft Foundry! ๐ฅ You can use it to build image and video editing apps. ๐จ๐ฒ SAM3 by @AIatMeta offers state of the art image AND video segmentation and understanding ๐ @alvarobartt made it available for easy deployment in Microsoft Foundry ๐ And created a step-by-step guide to walk you though it โคต๏ธ
New Keras release: 3.13 ๐ Some major new features: โข Model export to LiteRT (formerly TFLite) for mobile/edge โข GPTQ quantization support for post-training compression โข New Adaptive Pooling layers for dynamic architectures https://t.co/Ogmag7FYCY
Check out what's new: https://t.co/HtYruzv5P6
Tonight during the SLAMMYS be sure to tune into the @WWE app during #HeroinAllofUs category to vote for @NatbyNature https://t.co/9OGXxH7Vgc
Have you built your first cross-platform native app? NativeScript = Free + open source https://t.co/XHSDLu2ymh https://t.co/7yk2oAam9l
"Enter the sketch. It can be the first draft or the blueprint, a crude outline that will ultimately be ripped up or the architectural framework for a masterpiece." โ @passionweiss introduces #NativeSketches season two https://t.co/1BJO68mORW
ใๅใๆซ้ฒ็ฎใ ใไผๆฑใฉใคใใใใใคใฉในใใใใขใๅงใกใใ(ไปฎ)ใ ๅๅๅถไฝ๏ผใใใ #wf2019w #ใใคใใฃใ https://t.co/T80N5Bbyt6
SANEF has learnt that the EFF has denied access to journalists from Daily Maverick, Scorpio, amaBhungane, Rapport and Caxton publications, among several media outlets. https://t.co/mMsYGXCXiW
Getting into #NFTs Discussion panel with @avelykyy,@FirstNft, @Juscutum, @tacans_digital and @CoinindexAgency https://t.co/BkuNZr8AVi
๐ฏ ๐ฏ ๐ฏ Everything is Connected! ๐ฆฌโฆ READโฌ๏ธ Catch the Current! โ๐พ๐ค๐ฝโฆ https://t.co/dUqCuQc8wE
looking back at my best moments of this year and this day definitely tops it, @NatbyNature I donโt deserve you โค๏ธ https://t.co/XXn9wIHtFy
