VLMs show partial alignment with children's performance on six cognitive tasks, with stronger models matching better at task and item levels but struggling on matrix reasoning and mental rotation.
hub
Intuitive physics understanding emerges from self-supervised pretraining on natural videos.arXiv preprint arXiv:2502.11831
15 Pith papers cite this work. Polarity classification is still indexing.
15
Pith papers citing it
hub tools
citation-role summary
background 2
citation-polarity summary
roles
background 2polarities
background 2representative citing papers
YoCausal benchmark shows video diffusion models detect the arrow of time but lack genuine causal understanding relative to humans.
Multi-agent iterated learning produces emergent positionally disentangled communication protocols for latent physical properties from unsupervised video features.
A self-supervised framework learns implicit 3D physics by lifting V-JEPA features into voxels and performing volumetric feature advection conditioned on actions.
Asymmetric physics (high-fidelity non-diff simulator plus differentiable surrogates) enables end-to-end training of decentralized vision-based policies for up to 512 quadrupeds that transfer zero-shot to real hardware.
LaMo adds self-supervised latent motion priors via a motion drift loss during training and motion prior guidance during sampling to boost physical fidelity in video diffusion models like CogVideoX.
Latent prediction video models exhibit a distinct robustness profile across corruption, occlusion, fine-grained discrimination, and temporal sensitivity compared to other self-supervised video models when used as world models.
LeWM is a ~15M-parameter JEPA world model that trains end-to-end from pixels with only next-embedding prediction plus a Gaussian latent regularizer, cutting loss hyperparameters to one.
VisPhyWorld evaluates MLLMs' physical reasoning via executable code generation for video reconstruction, with VisPhyBench showing strong semantics but weak parameter inference and dynamics simulation.
Cambrian-S introduces VSI-SUPER benchmarks for long-horizon spatial recall and counting, shows data scaling yields 30% gains on existing tests, and demonstrates a self-supervised next-latent predictor using surprise outperforms baselines on the new spatial supersensing tasks.
Generative video models exhibit emergent zero-shot capabilities across perception, manipulation, and basic reasoning tasks.
Video foundation models encode intuitive physics knowledge that is strongest in V-JEPA at intermediate-to-late layers and depends on pretraining type and probe design.
The NTIRE 2026 Challenge released a public dataset of 2,000 videos with crowdsourced saliency maps and reported results from participating teams using standard quality metrics.
Phantom jointly models visual content and latent physical dynamics via a physics-aware video representation to generate physically consistent videos.
Standard success metrics for VLAs on complex chores overlook safety violations and intermediate failures, leading to exaggerated claims; new evaluation protocols are proposed to measure robustness and safety.
citing papers explorer
-
LEVANTE-bench: Multi-Scale Comparison of VLMs to Children Using Cognitive Tasks (or, "Is Your VLM Smarter Than a 5th Grader?")
VLMs show partial alignment with children's performance on six cognitive tasks, with stronger models matching better at task and item levels but struggling on matrix reasoning and mental rotation.
-
YoCausal: How Far is Video Generation from World Model? A Causality Perspective
YoCausal benchmark shows video diffusion models detect the arrow of time but lack genuine causal understanding relative to humans.
-
Emergent Compositional Communication for Latent World Properties
Multi-agent iterated learning produces emergent positionally disentangled communication protocols for latent physical properties from unsupervised video features.
-
Neural Voxel Dynamics: Learning Implicit 3D Physics via Volumetric Feature Advection
A self-supervised framework learns implicit 3D physics by lifting V-JEPA features into voxels and performing volumetric feature advection conditioned on actions.
-
Asymmetric physics enables efficient learning in quadrupedal robot swarms
Asymmetric physics (high-fidelity non-diff simulator plus differentiable surrogates) enables end-to-end training of decentralized vision-based policies for up to 512 quadrupeds that transfer zero-shot to real hardware.
-
LaMo: Self-Supervised Latent Motion Priors for Physical Realism in Video Generation
LaMo adds self-supervised latent motion priors via a motion drift loss during training and motion prior guidance during sampling to boost physical fidelity in video diffusion models like CogVideoX.
-
Latent Video Prediction Learns Better World Models
Latent prediction video models exhibit a distinct robustness profile across corruption, occlusion, fine-grained discrimination, and temporal sensitivity compared to other self-supervised video models when used as world models.
-
LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels
LeWM is a ~15M-parameter JEPA world model that trains end-to-end from pixels with only next-embedding prediction plus a Gaussian latent regularizer, cutting loss hyperparameters to one.
-
VisPhyWorld: Probing Physical Reasoning via Code-Driven Video Reconstruction
VisPhyWorld evaluates MLLMs' physical reasoning via executable code generation for video reconstruction, with VisPhyBench showing strong semantics but weak parameter inference and dynamics simulation.
-
Cambrian-S: Towards Spatial Supersensing in Video
Cambrian-S introduces VSI-SUPER benchmarks for long-horizon spatial recall and counting, shows data scaling yields 30% gains on existing tests, and demonstrates a self-supervised next-latent predictor using surprise outperforms baselines on the new spatial supersensing tasks.
-
Video models are zero-shot learners and reasoners
Generative video models exhibit emergent zero-shot capabilities across perception, manipulation, and basic reasoning tasks.
-
Do Video Foundation Models Understand Intuitive Physics? A Layerwise Probing Analysis
Video foundation models encode intuitive physics knowledge that is strongest in V-JEPA at intermediate-to-late layers and depends on pretraining type and probe design.
-
NTIRE 2026 Challenge on Video Saliency Prediction: Methods and Results
The NTIRE 2026 Challenge released a public dataset of 2,000 videos with crowdsourced saliency maps and reported results from participating teams using standard quality metrics.
-
Phantom: Physics-Infused Video Generation via Joint Modeling of Visual and Latent Physical Dynamics
Phantom jointly models visual content and latent physical dynamics via a physics-aware video representation to generate physically consistent videos.
-
How VLAs (Really) Work In Open-World Environments
Standard success metrics for VLAs on complex chores overlook safety violations and intermediate failures, leading to exaggerated claims; new evaluation protocols are proposed to measure robustness and safety.