Princeton Visual AI lab

Research on computer vision, AI/human interaction, vision+language, AI fairness+transparency. PI @orussakovsky. #PrincetonCS

[Princeton Visual AI lab @VisualAILab]

Can a few examples change what a vision-language model actually sees? Our new #NeurIPS2026 paper shows that in-context learning can substantially improve visual reasoning on counterfactual images, without updating model weights.

Check out @KunWang0129's thread below! 👇

[Kun Wang @KunWang0129]

Do vision-language models actually lack visual understanding, or are they held back by their learned priors?

In our #NeurIPS2026 paper, we find that in-context learning can help VLMs overcome misleading priors, improving accuracy by up to 27%.

🔗 https://t.co/aUQN9LbO1M

[Kun Wang @KunWang0129]

Do vision-language models actually lack visual understanding, or are they held back by their learned priors?

In our #NeurIPS2026 paper, we find that in-context learning can help VLMs overcome misleading priors, improving accuracy by up to 27%.

🔗 https://t.co/aUQN9LbO1M

[Princeton Visual AI lab @VisualAILab]

What happens to denoising when image generation takes just one forward pass? Turns out, the denoising process unfolds across the model's depth rather than over time. Check out @acaleb 's thread below! 👇

[Arnold Caleb Asiimwe @acaleb]

1/ I’m excited to share our work on “Depth as Time in One-Step Generative Models”!
https://t.co/M8OGqxykLk

TL;DR: We ask, when a model learns to generate an image in one forward pass, what happens to the denoising process? We observe that process unfolding over depth!

[William Yang @YangWilliam_]

I've always wondered: Where does denoising go in one-step generative models? Turns out, it's hidden in the layers!

A very surprising finding, and I'm super proud of Arnold for this work! 👏

[Arnold Caleb Asiimwe @acaleb]

1/ I’m excited to share our work on “Depth as Time in One-Step Generative Models”!
https://t.co/M8OGqxykLk

TL;DR: We ask, when a model learns to generate an image in one forward pass, what happens to the denoising process? We observe that process unfolding over depth!

[Princeton Visual AI lab @VisualAILab]

RT @cindy_x_wu: CFP: Continual World Models Workshop @ NeurIPS 2026

🗓 Submission deadline: Aug 29, 2026
📍 Sydney, Australia
🌐 https://t.co…

[Xindi Wu @cindy_x_wu]

So grateful for the outstanding honorable mention at the main conference for motion attribution as well as a best paper award at f2s workshop for our most recent work on memory for world models❤️

Both works ask a similar question: how can we better compress and control the temporal information that shapes a video model’s behavior.

What a week! I learned a lot from the community. Excited for what’s next in world models 🥰

#ICML2026

[Xindi Wu @cindy_x_wu]

So honored to receive the outstanding paper honorable mention for our motion attribution for video generation paper at #ICML2026

This wouldn’t happen without my amazing mentors and advisors! @paschalidoud_1 @JunGao33210520 Antonio Torralba @lealtaixe @orussakovsky @FidlerSanja @jonLorraine9 and many other friends and colleagues, this work would not have been possible without all of you and the huge support from @NVIDIAAI @PrincetonCS @MIT_CSAIL @UMichCSE @UofTCompSci @VectorInst 💪

[Xindi Wu @cindy_x_wu]

Excited to share several recent works at #ICML2026! 🇰🇷 Happy to meet & chat about world models, video generation & data centric multimodal ML! DMs are open

Motion Attribution for Video Generation will be presented as an oral + poster:
- Tue Jul 7, 10:30–10:45 AM KST📍Auditorium
- Tue Jul 7, 2:00–3:45 PM KST📍Hall A #2807
https://t.co/TbKXjQMN3H

Also presenting two workshop works:
Addressable Memory for Video World Models at the From Frames to Stories (F2S) workshop
- Fri Jul 10, 14:10–14:25 KST📍Room 402
https://t.co/6mmp5jKbft

Variance Reduction for Expectations with Diffusion Teachers at the SPIGM workshop, led by amazing co-author Jesse Bettencourt @jessebett
- Fri Jul 10, 11:40 AM–1:30 PM KST📍Hall D1
https://t.co/AvAsXkNk2b

https://t.co/TbKXjQMN3H

[Princeton Computer Science @PrincetonCS]

At the end of May, David Beeson ‘26, graduated from @Princeton with a degree in computer science. In the fall, he will enroll at the @UniofOxford to pursue a master’s degree in computer science with a specialization in artificial intelligence.

[Ye Zhu @szyezhu]

How can we distill large text-to-image models without losing the diversity and controllability of the original teacher model?
Excited to share our new ICML 2026 paper: Restoring Initial Noise Sensitivity in Text-to-Image Distillation via Geometric Alignment.
#icml #generation
Fast T2I distillation often maintains high image quality but makes the student much less sensitive to initial noise. This hurts an important property of generative models: changing the input noise should meaningfully change the output, which matters for diversity, layout variation, and noise-driven control. In this work, we trace this issue to standard distillation objectives that enforce pointwise output matching and unintentionally flatten the teacher’s local input-output geometry. We propose GAD (Geometry-Aware Distillation), a plug-and-play objective that preserves the teacher’s local functional behavior by aligning responses to noise perturbations through Jacobian-vector products.
✅ Better diversity
✅ Stronger layout/control transfer
✅ Preserved visual fidelity
✅ No extra inference cost

Paper: https://t.co/sBt1xSm1FE
Code: https://t.co/bmE8qX5cNE

[Princeton Artificial Intelligence @PrincetonAInews]

Here's a roundup of the work that will be showcased by Princeton researchers this week at #CVPR2026
https://t.co/KUofXEQQSQ

[William Yang @YangWilliam_]

I will be presenting BeyondObjects this week at #CVPR2026. Come check out our poster on Thursday 4:30-5:30pm at the syndata4cv workshop (Room 607) or Saturday 4:45-6:45pm at ExHall A & F poster location 99. Happy to chat!

[Xindi Wu @cindy_x_wu]

Happy to share that our paper “Motion Attribution for Video Generation” has been accepted to #ICML2026 as a Spotlight (top 2.2%) 🎉

https://t.co/TbKXjQMfe9

TL;DR: We introduce MOTIVE, a scalable, motion-centric data attribution framework for video generation that identifies which training clips improve or degrade motion dynamics, enabling better data curation and beyond.

[Jonathan Williams @joncopewilliams]

Excited to share that RLTT has been accepted to #ICML2026! Looking forward to connecting with others working on Looped Transformers/Latent Reasoning/Mathematical Reasoning in Seoul🇰🇷!

[Xindi Wu @cindy_x_wu]

Honored to receive the Best Paper Award at the Test-Time Updates Workshop @ ICLR 2026 🏆

Huge thanks to coauthors and workshop organizers!
https://t.co/585kRh5NzF

[Xindi Wu @cindy_x_wu]

Honored to receive the 2026 Apple Scholars in AIML PhD fellowship! 🍎

Extremely grateful to my advisor @orussakovsky and all the incredible mentors, collaborators and friends I’ve had throughout the journey. Excited to push toward more scalable and capable multimodal system!

https://t.co/WhGTaD2PDa

[Tyler Zhu @tyleryzhu]

We'll be presenting this work on video-text alignment this week at #ICLR2026!

Come check out our poster in Pavillion 4 #3405 in the Thursday AM session from 10:30am to 1pm. Happy to chat about PRH, repr learning, video understanding, or how vision plays into text models!

[Allison Chen @allisonchen_227]

How should we talk about LLMs? Does it matter if we frame them as a machines 📠, tools ⚒️, or companions 👥? In our #CHI2026 paper, that these framings can alter what people believe about LLMs and how they use them. See 🧵for more!

[Kaleb Newman @kalebnewman8]

Video models surprisingly can solve mazes, but inconsistently. We understand little about how they reason, making it hard to use such abilities.

We investigate the denoising process and find models commit to a plan early, letting us screen far more candidates for better perf.
🧵