Video wird geladen...
Video konnte nicht geladen werden
⚡️ 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 Aufrufe • vor 11 Monaten •via X (Twitter)
4 Kommentare

Anirudha Majumdarvor 11 Monaten
❌ 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 Majumdarvor 11 Monaten
Check out the paper and the project website to find out more! 📃 Paper: 🌐 Website:

Decoded Institutevor 11 Monaten
This is the missing layer: not just generating, but calibrating what the model doesn’t know. Without uncertainty, ‘plausible’ becomes a liability.

Himanshu Kumarvor 11 Monaten
Uncertainty quantification in action-conditioned models requires evaluating calibration metrics; the ECE score is 0.12.
