RPent
RPent is an open framework for building embodied agents that combine perception, reasoning, memory and execution into one physical agent, meant to keep acquiring new capabilities through continued interaction with the real world.
RPent, short for Recursive Physical Agent, is an open framework for building embodied agents that evolve through recursive interaction with the physical world. Rather than relying on one fixed foundation model, it combines heterogeneous components for perception, reasoning, memory, execution and self-evolution into a single physical agent, built on a service-oriented, standardized and composable architecture.
The project reports results on benchmarks including LIBERO, LIBERO-PRO, RoboCasa365 and RoboTwin, and describes four intended audiences: embodied-intelligence researchers targeting long-horizon manipulation, online and reinforcement-learning researchers studying self-evolving agents, robotics developers deploying solutions on real hardware, and end users who install RPent with real-robot extensions to run predefined tasks without machine-learning expertise.