RoboGaze presents a structured multi-agent VLM pipeline and robotics-specific error taxonomy that improves video evaluation metrics by up to 43 F1 points over zero-shot baselines on a 382-clip dataset.
hub Canonical reference
Rethinking video generation model for the embodied world
Canonical reference. 80% of citing Pith papers cite this work as background.
16
Pith papers citing it
Background
80% of classified citations
hub tools
citation-role summary
background 4
other 1
citation-polarity summary
years
2026 16representative citing papers
Processed egocentric human video outperforms teleoperated real-robot trajectories as pretraining data for embodied foundation models, delivering 24% lower validation loss and 52.5-90% higher task success rates under matched post-training protocols.
Dream.exe evaluates 8 video generation models on 101 manipulation tasks by converting generated videos into executable robot trajectories in a simulator, finding measurable success rates that visual metrics do not predict.
A2A flow matching starts action generation from prior proprioceptive actions in latent space to enable single-step high-quality predictions in robotic policies.
Agentic VLM rewards plus dynamic-aware GRPO rollouts let embodied world models explore more diversely while cutting reward hacking and lifting domain scores on Cosmos-Predict2.5 and Kairos3.0-Robot.
HumanNet is a 1M-hour human-centric video dataset with interaction annotations that enables better vision-language-action model performance than equivalent robot data in a controlled test.
ExoActor uses exocentric video generation to implicitly model robot-environment-object interactions and converts the resulting videos into task-conditioned humanoid control sequences.
Veo-3 video predictions enable approximate task-level robot trajectories in zero-shot settings but require hierarchical integration with low-level VLA policies for reliable manipulation performance.
PhysisForcing applies trajectory and relational alignment losses to DiT features in video models, improving physical plausibility on R-Bench, PAI-Bench, and EZS-Bench while raising closed-loop robotic success rates from 16% to 24%.
The paper proposes an L0-L7 evidential ladder for evaluating world models in embodied decision-making, prioritizing interventional action fidelity and policy optimization utility over visual plausibility.
WorldOlympiad is a new benchmark decomposing world-model evaluation into physical, geometry, and interaction tracks using segmentation, MLLM judges, Gaussian splatting, and action prompts on diverse scenarios.
StableVLA adds an Information Bottleneck Adapter to VLA models that improves robustness to visual corruptions by 30% on average with under 10M extra parameters and no extra data, even when using a much smaller backbone.
In twisted bilayer nodal d-wave superconductors, interlayer hopping creates nodes on the C2 axis and Bogoliubov flat bands when the single-layer Berry connection is parallel to that axis.
Using VGGT's reconstruction error as a self-supervised reward, DPO post-training with ~2,500 preference pairs improves 3D consistency of CogVideoX-based video generation while preserving perceptual quality.
A comprehensive survey that organizes the literature on world models in robot learning, their roles in policy learning, planning, simulation, and video-based generation, with connections to navigation, driving, datasets, and benchmarks.
citing papers explorer
-
RoboGaze: Evaluating Robot World Models via Structured Vision-Language Analysis
RoboGaze presents a structured multi-agent VLM pipeline and robotics-specific error taxonomy that improves video evaluation metrics by up to 43 F1 points over zero-shot baselines on a 382-clip dataset.
-
HumanScale: Egocentric Human Video Can Outperform Real-Robot Data for Embodied Pretraining
Processed egocentric human video outperforms teleoperated real-robot trajectories as pretraining data for embodied foundation models, delivering 24% lower validation loss and 52.5-90% higher task success rates under matched post-training protocols.
-
Dream.exe: Can Video Generation Models Dream Executable Robot Manipulation?
Dream.exe evaluates 8 video generation models on 101 manipulation tasks by converting generated videos into executable robot trajectories in a simulator, finding measurable success rates that visual metrics do not predict.
-
Action-to-Action Flow Matching
A2A flow matching starts action generation from prior proprioceptive actions in latent space to enable single-step high-quality predictions in robotic policies.
-
Reward as An Agent for Embodied World Models
Agentic VLM rewards plus dynamic-aware GRPO rollouts let embodied world models explore more diversely while cutting reward hacking and lifting domain scores on Cosmos-Predict2.5 and Kairos3.0-Robot.
-
HumanNet: Scaling Human-centric Video Learning to One Million Hours
HumanNet is a 1M-hour human-centric video dataset with interaction annotations that enables better vision-language-action model performance than equivalent robot data in a controlled test.
-
ExoActor: Exocentric Video Generation as Generalizable Interactive Humanoid Control
ExoActor uses exocentric video generation to implicitly model robot-environment-object interactions and converts the resulting videos into task-conditioned humanoid control sequences.
-
Veo-Act: How Far Can Frontier Video Models Advance Generalizable Robot Manipulation?
Veo-3 video predictions enable approximate task-level robot trajectories in zero-shot settings but require hierarchical integration with low-level VLA policies for reliable manipulation performance.
-
PhysisForcing: Physics Reinforced World Simulator for Robotic Manipulation
PhysisForcing applies trajectory and relational alignment losses to DiT features in video models, improving physical plausibility on R-Bench, PAI-Bench, and EZS-Bench while raising closed-loop robotic success rates from 16% to 24%.
-
How Should World Models Be Evaluated for Embodied Decision-Making? A Decision-Making-Centric Position
The paper proposes an L0-L7 evidential ladder for evaluating world models in embodied decision-making, prioritizing interventional action fidelity and policy optimization utility over visual plausibility.
-
WorldOlympiad: Can Your World Model Survive a Triathlon?
WorldOlympiad is a new benchmark decomposing world-model evaluation into physical, geometry, and interaction tracks using segmentation, MLLM judges, Gaussian splatting, and action prompts on diverse scenarios.
-
StableVLA: Towards Robust Vision-Language-Action Models without Extra Data
StableVLA adds an Information Bottleneck Adapter to VLA models that improves robustness to visual corruptions by 30% on average with under 10M extra parameters and no extra data, even when using a much smaller backbone.
-
Video Generation Models as World Models: Efficient Paradigms, Architectures and Algorithms
In twisted bilayer nodal d-wave superconductors, interlayer hopping creates nodes on the C2 axis and Bogoliubov flat bands when the single-layer Berry connection is parallel to that axis.
-
VideoGPA: Distilling Geometry Priors for 3D-Consistent Video Generation
Using VGGT's reconstruction error as a self-supervised reward, DPO post-training with ~2,500 preference pairs improves 3D consistency of CogVideoX-based video generation while preserving perceptual quality.
-
World Model for Robot Learning: A Comprehensive Survey
A comprehensive survey that organizes the literature on world models in robot learning, their roles in policy learning, planning, simulation, and video-based generation, with connections to navigation, driving, datasets, and benchmarks.
- ROBOSHACKLES: A Safety Dataset for Human-Injury Prevention in Embodied Foundation Models