(1/N) Are you looking for #Vision #Transformer papers in various areas? Check out this list of papers including a broad range of different tasks! github.com/cmhungsteve/Aweso… Feel free to share with others😀 @Montreal_AI @machinelearnflx @hardmaru @ak92501 @arankomatsuzaki @omarsar0

SpatialClaw NVIDIA drops a training-free spatial reasoning agent that uses code as its action interface. A VLM writes Python in a persistent kernel, composes perception tools, inspects results, and revises its plan—no fine-tuning needed. 11.2 points over prior agents on 20 benchmarks.

We are presenting Fast-ThinkAct at 10:45am this morning at Poster #469. Welcome to drop by and discuss!

As NVIDIA pushed its first ever open weights autopilot model alpamayo-R1 A key frontier is native 4D understanding: not just “describe the video” but reason about depth motion 3D interactions over time, even for a specific region. This paper introduces 4D-RGPT (also from NVIDIA!), which tackles this by distilling 4D perception (depth motion cues) from a frozen expert into an MLLM during training, using both latent-feature and explicit-signal distillation, plus timestamp positional encodings, with training-only modules (so no extra inference cost). On a new benchmark they developed for region-level undestanding, R4D-Bench, this approach achieves 4.3% across the bench! And tops non-region based 3D & 4D benchmarks by 5.3%

NVIDIA just released 4D-RGPT on Hugging Face A CVPR 2026 Highlight model for region-level 4D video understanding. It learns depth and motion from an expert at training time — with no extra cost at inference.

T4V @CVPR is starting!!

Come to the T4V Workshop next week at CVPR for the latest developments in Vision Transformers and Multimodal AI!

T4V @CVPR starts today🚀🚀 1:45–5:40pm, Room 607 Join talks/discussion on Transformers for image, video, 3D, MLLMs, efficient attention & SSMs/Mamba

What comes after today’s visual backbones? At T4V @CVPR 2026, we’re bringing the community together for a focused half-day workshop on Transformers for Vision and Multimodal AI — covering image, video, 3D, MLLMs, efficient attention, SSMs/Mamba, and the next generation of visual architectures. 📍 Wed June 3 · Room 607 🕐 1:45–5:40 pm (Denver local time) Invited speakers: @RanjayKrishna, @thoma_gu, @sherryyangML, @jcniebles, @liuzhuang1234, and @TongPetersb. Join us at CVPR: sites.google.com/view/t4v-cv… @UNC @NVIDIAAI @NVIDIAAIDev @AIatMeta @ImagineEnpc @NJU1902 @BaskinEng #CVPR2026 #T4V #MultimodalAI #NVIDIA

We will be presenting V2V-Got and V2V-LLM at the 🇦🇹 #ICRA2026 conference and the 🇺🇸 #CVPR2026 workshop next week! 🚗🤖🚗 We explore the intersection of Cooperative Autonomous Driving and Multimodal LLMs enabling vehicles to perform human-like physical reasoning. ⬇️Threads

Thank you for all the interest in illoca Tracing Paper. The previous video was about why we built it. This one shows what architects can do with it: Sketch. Mark up. Describe ideas in text, or show them with references. Turn intent into editable 2D and 3D designs, so architects can spend less time rebuilding and more time designing.

EoRA is a great and useful code & work done by @nbasyl_tw, @CMHungSteven and the @nvidia teams. (CUDA kernel supported)