Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

⚡️ Introducing video world models that know when they don't know! 📽️ Action-conditioned video models can hallucinate a lot, generating videos that look plausible but are incorrect! How can we rigorously quantify their uncertainty, and know when to trust generations?

12,038 görüntüleme • 11 ay önce •via X (Twitter)

4 Yorum

Anirudha Majumdar profil fotoğrafı
Anirudha Majumdar11 ay önce

❌ Hallucinations hinder trustworthy video generation, e.g., for scalable evaluation of generalist robot policies and visual planning! 🚀 Using proper scoring rules, our approach generates dense confidence predictions at the subpatch level that are physically interpretable and aligned with observations.

Anirudha Majumdar profil fotoğrafı
Anirudha Majumdar11 ay önce

Check out the paper and the project website to find out more! 📃 Paper: 🌐 Website:

Decoded Institute profil fotoğrafı
Decoded Institute11 ay önce

This is the missing layer: not just generating, but calibrating what the model doesn’t know. Without uncertainty, ‘plausible’ becomes a liability.

Himanshu Kumar profil fotoğrafı
Himanshu Kumar11 ay önce

Uncertainty quantification in action-conditioned models requires evaluating calibration metrics; the ECE score is 0.12.

Benzer Videolar