Claims Browser

Universal physical laws govern spatiotemporal dynamics regardless of the actor

"CLAP is grounded in the insight that universal physical laws govern spatiotemporal dynamics regardless of the actor."

Transition realization is a consequential architectural choice in latent World Action Models

"These results establish transition realization as a consequential architectural choice in latent WAMs."

Advanced plant characterization requires the integration of active physiological sensing modalities such as chlorophyll fluorescence beyond image-based measurements

"advanced plant characterization requires the integration of active physiological sensing modali- ties such as chlorophyll fluorescence"

The proposed robotics system opens new directions for autonomous agricultural inspection and plant-aware manipulation

"the proposed system provides a robotics-driven approach to high-resolution plant phenotyping and opens new directions for autonomous agricultural inspection and plant-aware manipulation"

The intrinsic causality of reservoir topology provides a robust framework for autonomous reliability and fault-tolerant physical intelligence

"The results suggest that the intrinsic causality of the reservoir's topology provides a robust framework for autonomous reliability and fault-tolerant physical intelligence."

Gemini Robotics ER 2 represents a step change in video understanding, tool orchestration, and multi-robot collaboration for robotic applications

"It represents a step change in video understanding, tool orchestration, and multi-robot collaboration for robotic applications."

Spatial intelligence involves not just perceiving and generating worlds, but also interacting with them

Vision-Only Long-Horizon Navigation (VoLN) is a better formulation where route-relevant information comes from locally observable in-scene cues rather than external instructions

Test time training is a promising approach for robotic learning

Rational risk/reward computations do not apply to self-driving car incidents and things can go south very quickly for a self-driving company

The VLA bottleneck stems from conflating physical competence learning and semantic alignment, where only the latter requires language supervision

Simply increasing number of viewpoints is insufficient for spatial generalization in VLA models

Vision-Language Navigation research has emphasized high-level reasoning while low-level action representation remains underexplored

Robot self-improvement can be achieved by building capabilities one skill at a time

ASPIRE gives new meaning to sim2real transfer and cross-embodiment transfer by using coding agents that observe multimodal sensory traces and launch evolutionary search over control programs

Current examples of autonomous robot experimentation loops are suggestive at best, with significant limitations remaining for real-world deployment