The workshop schedule is as follows (all times in the PST timezone, i.e., local time in Vancouver).

Date: December 14, 2024

Room: West Meeting Room 202-204

NeurIPS link: https://neurips.cc/virtual/2024/workshop/84703 (requires a NeurIPS registration for live stream)

All papers will be presented in poster sessions.

Note: slides of invited talks will be made available here soon.



09:00-09:15 Opening Remarks
09:15-09:45 Invited Talk: Sherry Yang (Google DeepMind)
Self-Supervised World Modeling from Internet Data [slides]
09:45-10:15 Invited Talk: Pauline Luc (Google DeepMind)
Self-supervision for General Video Understanding Beyond Semantics [slides]
10:15-10:30 Coffee Break
10:30-11:00 Invited Talk: Hanna Hajishirzi (University of Washington & AI2)
OLMo & Molmo: Open Textual and Visual Language Models [slides]
11:00-11:30 Hilde Kuehne (Univ. of Tuebingen & MIT-IBM Watson AI Lab)
Advances in Self-supervised Multimodal Learning [slides]
11:30-12:00 Oral Talks:
  • In-Context Symmetries: Self-Supervised Learning through Contextual World Models.
  • A Unifying Framework for Action-Conditional Self-Predictive Reinforcement Learning.
12:00-12:30 Lunch Break
12:30-13:50 Poster Session
14:00-14:30 Invited Talk: Trevor Darrell (UC Berkeley)
From Unsupervised Segmentation to Visual Prompting [slides]
14:30-15:00 Invited Talk: Alan Yuille (Johns Hopkins University)
Supervision of 3D-aware models by Synthetic Data [slides]
15:00-15:30 Invited Talk: Phillip Isola (MIT)
Representation Learning from Human Feedback [slides]
15:30-16:00 Invited Talk: Lili Yu (FAIR, Meta)
Paths Towards Deep Fused Multimodal Modeling [slides]
4:00-4:30 Invited Talk: Ziwei Liu (Nanyang Technological University)
From High-fidelity 3D Generative Models to Dynamic Embodied Learning [slides]
16:30-17:00 Oral Talks:
  • Occam's Razor for Self Supervised Learning: What is Sufficient to Learn Good Representations?
  • Neural Embedding Ranks: Aligning 3D latent dynamics with movement for long-term decoding.