Newest papers

AgenticRoboRL: Bringing Coding-Agent Post-Training to Agentic Robotics

Xinxin Liu, Bingxin Xu, Wenrui Bao, Ming Li, Wentao Bao, Chen Chen, Yuzhang Shang · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.28589 · cited by 0 works (OpenAlex) · open-access PDF

German

Agentische Robotersysteme verbinden Planer mit Roboterarmaturen, um komplexe Framework-Manipulationsaufgaben auszuführen. Die jüngste Arbeit verbessert diese Systeme durch bessere Armaturen und externes Gedächtnis, beruht jedoch auf leistungsstarken geschlossenen Quellmodellen als Planer und verbessert nicht die eigenen Fähigkeiten der Planer. Im Gegensatz dazu haben Codierungsagenten AgenticRoboRL erheblich verbessert, indem sie das Modell selbst nach dem Training ausführen. Daher untersuchen wir, ob das Nach-Training auch Roboterplaner verbessern kann. Jedoch erfordert die Anwendung über Roboterarmaturen hinweg die Anpassung verschiedener Ausführungsmechanismen, einschließlich der VLA-Fähigkeiten-Invocation und der Steuerungs-Code-Gener

Machine translation (Meta M2M100); the English original follows.

Abstract

Agentic robotics systems combine planners with robot harnesses to perform complex manipulation tasks. Recent work improves these systems through better harnesses and external memory, but relies on powerful closed-source models as planners and does not improve the planners' own capabilities. In contrast, coding agents have improved substantially by post-training the model itself. We therefore investigate whether post-training can also improve robot planners. However, applying it across robot harnesses requires accommodating different execution mechanisms, including VLA skill invocation and control-code generation. To this end, we introduce AgenticRoboRL, a framework for post-training robot planners across different harnesses. The training recipe combines supervised fine-tuning with multi-turn reinforcement learning. To accommodate different execution mechanisms, harness-specific adapters connect recorded planner interactions with task outcomes, enabling a common RL training pipeline while keeping execution components fixed. On LIBERO-PRO, AgenticRoboRL raises the success rate of a 9B-parameter planner from 18.1% to 71.0% without external memory, approaching the 72.1% achieved by Codex (GPT-5.5) with external memory. These results demonstrate post-training as an effective approach for improving robot planners. We will release the framework and trained planner checkpoints.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Robot, Robotics, Planner, Artificial intelligence, Computer science, Task (project management)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange