Primary-source guide

Research archive

Landmark Physical AI papers with plain-language explanations, prerequisites, impact, and direct primary-source links.

PaperOrganizationYearConcepts
NVIDIA Isaac GR00T N1.7GR00T N1.7 is an open vision-language-action model for generalized humanoid manipulation skills.
NVIDIA2026VLAHumanoidsDeployment
Gemini Robotics-ER 1.6Gemini Robotics-ER 1.6 improves spatial reasoning, multi-view understanding, tool use, and safety-oriented robot reasoning.
Google DeepMind2026Embodied ReasoningSafetyPlanning
Genie 3: A General-Purpose World ModelGenie 3 generates interactive environments that can be explored in real time from text descriptions.
Google DeepMind2026World ModelsInteractive Environments
NVIDIA Isaac GR00T-DreamsGR00T-Dreams uses world foundation models to generate synthetic robot trajectories from a single image and instruction.
NVIDIA2025Synthetic DataRobot LearningWorld Models
Gemini Robotics 1.5Gemini Robotics 1.5 turns visual observations and instructions into motor commands while supporting multi-step physical tasks.
Google DeepMind2025VLAEmbodied ReasoningRobot Control
Marble: A Multimodal World ModelMarble generates persistent 3D worlds from text, images, video, panoramas, or coarse 3D layouts.
World Labs2025Spatial IntelligenceWorld Models3D
Aria Gen 2 Pilot DatasetAria Gen 2 Pilot Dataset captures daily activities with research glasses and synchronized multimodal sensor data.
Meta Reality Labs2025Egocentric DataWearablesRobotics
Figure HelixHelix is Figure's generalist humanoid VLA model for onboard perception, reasoning, and full-upper-body control.
Figure AI2025VLAHumanoidsDexterity
DROID: Distributed Robot Interaction DatasetDROID provides diverse in-the-wild robot manipulation demonstrations across many scenes, tasks, and collectors.
Stanford, Berkeley, Google DeepMind and collaborators2024DatasetsManipulationRobot Learning
Open X-Embodiment and RT-XOpen X-Embodiment combines robot datasets across institutions and trains policies that transfer across embodiments.
Google DeepMind and 20+ institutions2023DatasetsCross-EmbodimentVLA
DreamerV2DreamerV2 extends latent imagination with discrete representations and reaches human-level Atari performance.
Google Research2020World ModelsAtari
World ModelsA compact latent model can let an agent learn behavior inside its own predicted environment.
Google Brain / IDSIA2018World ModelsReinforcement Learning