๐ฌ Excited to share our paper: Joints, Camera, Action! (JCA)
JCA generates video & 3D motion together for any rig!
Video models generate motion you can watch. But what if you could use that motion to animate a 3D asset, with motion that responds to its surroundings?
https://t.co/Y7smggHv4w
๐งต
[1/6] State-of-the-art diffusion LMs decode in blocks: each block is refined and self-edited before the next one starts.
But a block is only as fast as its slowest token.
We introduce B3: Blurring the Block Boundaries ๐งต https://t.co/phKLaRLxN5
Minimax h3 ๐ฒ Motion Time to move
implementation of time-to-move dual clock denoising for Minimax H3, allowing for motion control by dragging objects into motion, or mask part of a video
and change what happens.
github:
https://t.co/KK4OB9edGt
workflow:
https://t.co/2f8047N6sH
Meet ReLift, our latest work on 3D tracking! ๐๏ธ
Rethinking 3D tracking with the right combination:
Strong 2D trackers
๏ผ 3D geometry
๏ผ Efficient trajectory corrector
โโโโโโโโโโโโโโ
ใ Dense 3D tracking,
ใ even through occlusions!
๐ Introducing ReLift: dense 3D tracking from monocular video.
Strong 2D trackers provide correspondences.
Geometry models produce depth & cameras.
We show how to combine them, repair 3D errors, and recover occluded motion for SOTA 3D tracking.
All it takes: one small corrector + off-the-shelf models.
๐ Project: https://t.co/9Xe5hiGSsX
@MarounKhriesh@NoamRot@yoni_kasten@orlitany
Coding agents can build an app. What if they could give it a body?
Meet Metatron: a versatile System 1 controller for humanoids. A goal-based interface connects System 2 planners, including coding agents, to whole-body movement.
A few examples below. 1/5
1/4 We trained an 8B safety classifier on just 1,275 carefully annotated prompts.
It beat GPT-5.4, Claude Sonnet 4.6, and leading open-weight guardrails trained on far more data, on average across 5 safety benchmarks.
Quality > quantity.
๐ฌ Excited to share our paper: Joints, Camera, Action! (JCA)
JCA generates video & 3D motion together for any rig!
Video models generate motion you can watch. But what if you could use that motion to animate a 3D asset, with motion that responds to its surroundings?
https://t.co/Y7smggHv4w
๐งต
We build ~200K videoโmotion pairs across diverse rigs.
A major part is โObjaverse in Videoโ: a video editor places existing animated asset renders in suitable scenes, creating data at scale for arbitrary rigs. We add human and robot data.
Do vision-language models look at the image regions they describe? In our new paper, we present Self-Saliency, a method for aligning visual attention with reasoning text.
https://t.co/WyqXrvGYPc w/ @GalChechik@DalalGal
Excited to share our work, โToken-to-Token Alignment of Text Embeddings for Semantic Blending,โ accepted as a NeurIPS 2026 Spotlight ๐
Can we directly interpolate between semantic concepts?
Text embedding spaces already encode a continuous semantic structure, the missing piece is alignment.
Once aligned, simple interpolation becomes meaningful, naturally enabling multiple applications.
Joint work with @MokadyRon@OPatashnik@DanielCohenOr1