The Flow has not summarised this story yet — read it at arXiv Computation and Language.
OccJEPA treats occlusion as masking for amodal video segmentation by dropping tokens outside an object's visible mask and using a pre‑trained V‑JEPA 2.1 predictor to fill in the missing features, followed by a lightweight head that decodes these into complete masks. The resulting deterministic model is 33.5× faster than diffusion‑based state‑of‑the‑art methods and achieves higher accuracy on SAIL‑VOS, TAO‑Amodal (including a held‑out test split), and Kubric. Experiments show that the performance gains stem from the predictor’s pre‑trained weights rather than its size, with models ranging from 87 M to 1.85 B parameters.
arXiv:2609.37378v1 Announce Type: cross Abstract: Latent world models are trained to predict what happens next, so nothing in their objective separates what an action caused from what merely co-occur...
arXiv:2609.09528v1 Announce Type: new Abstract: Video large language models (Video-LLMs) are increasingly used as the perceptual front end of world models, a role that assumes they can read motion: h...
LeVJEPA is a video encoder that eliminates the need for architectural asymmetries, exponential-moving-average target encoders, stop-gradients, and capacity-limited predictors used in prior self‑supervised methods. It trains a single encoder with an invariance loss over global and local views, regularized by SIGReg to prevent collapse, and achieves strong performance with far less pretraining compute. The approach also allows block‑causal attention, making temporal ordering a property of the encoder itself, and matches or surpasses state‑of‑the‑art baselines on both appearance‑centric and motion‑centric benchmarks.
CALIPER is a new benchmark that tests whether pretrained visual encoders can infer physical properties such as mass and friction from images. The test involves striking an object twice at known speeds, showing a third strike only up to contact, and asking a linear readout on frozen features to predict how far the object slides. Results show that in clean, fixed‑camera scenes all representations perform similarly, but when camera, lighting, and clutter are varied, only encoders that truly infer physics—like V‑JEPA 2—maintain performance, while random or raw pixel representations fail.
arXiv:2608.30692v1 Announce Type: new Abstract: Video world models are increasingly used as simulators, yet visual fidelity alone does not show that a model maintains the hidden state of the world. W...