RACaP: Agentic Reasoning, Acting, and Coding as Policies for Evolvable Robot Learning

Zexi Li; Yehang Zhang; Haojian Huang; Bohan Zhou; Wenqian Li; Chenxu Wang; Yifan Chang; Yangkai Wei; Tianyi Zhang; Ying-Cong Chen; Kaiwen Zhou; Yinchuan Li; James Cheng

arXiv · 2609.29394

RACaP: Agentic Reasoning, Acting, and Coding as Policies for Evolvable Robot Learning

Zexi Li·Yehang Zhang·Haojian Huang·Bohan Zhou·Wenqian Li·Chenxu Wang·Yifan Chang·Yangkai Wei·Tianyi Zhang·Ying-Cong Chen·Kaiwen Zhou·Yinchuan Li·James Cheng

View original source ↗Read source document ↗Export metadata ↓

Abstract

General-purpose robot agents must learn from experience, transfer to new tasks, and act efficiently. Code as Policies (CaP) methods generate and repair programs at runtime, incurring latency and entangling reusable mechanisms with task-specific decisions. We introduce RACaP, an agentic framework that moves coding to evolution and uses a Reasoning-and-Acting (ReAct) loop to call frozen, typed Policy APIs at deployment. A two-phase strategy combines capability curriculum learning with autonomous self-evolution to improve the APIs, the ReAct harness, and experience memory. The APIs encode reusable physical mechanisms while exposing arguments for runtime adaptation. ReAct combines task-specific working memory, long-term experience memory, and visual feedback to select actions, verify outcomes, and recover from failures without modifying source code. RACaP achieves 54.4% success on LIBERO-90, 45.0% on zero-shot LIBERO-PRO, and 46.0% on LIBERO-Long, compared with at most 4.0% for CaP baselines on long-horizon tasks. On LIBERO-PRO, it achieves 2.5 times the success rate of CaP baselines and a 1.9-fold speedup in median policy time. For efficient on-robot deployment, rejection-sampled fine-tuning distills GPT-5.6 ReAct decisions into Qwen3-VL-8B-Instruct, yielding a 13.2-fold per-decision inference speedup and reducing repeated physical calls from 16 to 4. These results show that separating reusable code from runtime decisions supports continued evolution, effective transfer, and efficient long-horizon control.

Explore related subjects

cs.RO

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Zexi Li, Yehang Zhang, Haojian Huang, Bohan Zhou, Wenqian Li, Chenxu Wang, Yifan Chang, Yangkai Wei, Tianyi Zhang, Ying-Cong Chen, Kaiwen Zhou, Yinchuan Li, James Cheng. 2026-09-24. RACaP: Agentic Reasoning, Acting, and Coding as Policies for Evolvable Robot Learning. https://arxiv.org/abs/2609.29394

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Physics-Guided Residual Reinforcement Learning for Humanoid Narrow-Path Traversal

Traversing narrow paths is challenging for humanoid robots due to the sparse and safety-critical footholds required. Purely template-based or end-to-end reinforcement learning-based methods suffer from such harsh terrains. This paper proposes a two stage training framework for such narrow path traversing tasks, coupling a template-based foothold planner with a low-level foothold tracker from Stage-I training and a lightweight perception aided foothold modifier from Stage-II training. With the curriculum setup from flat ground to narrow paths across stages, the resulted controller in turn learns to robustly track and safely modify foothold targets to ensure precise foot placement over narrow paths. This framework preserves the interpretability from the physics-based template and takes advantage of the generalization capability from reinforcement learning, resulting in easy sim-to-real transfer. The learned policies outperform purely template-based or reinforcement learning-based baselines in terms of success rate, centerline adherence and safety margins. Validation on a Unitree G1 humanoid robot yields successful traversal of a 0.2m wide and 3m long beam for 20 trials without any failure.

cs.RO↗

arXiv2026

StageCraft: Execution Aware Mitigation of Distractor and Obstruction Failures in VLA Models

Large scale pre-training on text and image data along with diverse robot demonstrations has helped Vision Language Action models (VLAs) to generalize to novel tasks, objects and scenes. However, these models are still susceptible to failure in the presence of execution-time impediments such as distractors and physical obstructions in the robot's workspace. Existing policy improvement methods finetune base VLAs to improve generalization, yet they still struggle in unseen distractor settings. To address this problem, we investigate whether internet-scale pretraining of large vision-language models (VLMs) can be leveraged to reason about these impediments and mitigate policy failures. To this end, we propose StageCraft, a training-free approach to improve pretrained VLA policy performance by manipulating the environment's initial state using VLM-based in-context reasoning. StageCraft takes policy rollout videos and success labels as input and leverages VLM's reasoning ability to infer which objects in the initial state need to be manipulated to avoid anticipated execution failures. StageCraft is an extensible plug-and-play module that does not introduce additional constraints on the underlying policy, and only requires a few policy rollouts to work. We evaluate performance of state-of-the-art VLA models with StageCraft and show an absolute 40% performance improvement across three real world task domains involving diverse distractors and obstructions. Our simulation experiments in RLBench empirically show that StageCraft tailors its extent of intervention based on the strength of the underlying policy and improves its performance with more in-context samples. Videos of StageCraft in effect can be found at https://stagecraft-decorator.github.io/stagecraft/ .

cs.RO↗

arXiv2026

RHINO-AR: An Augmented Reality Exhibit for Teaching Mobile Robotics Concepts in Museums

Understanding autonomous mobile robots is challenging for people without robotics expertise, because key processes such as sensing, mapping, and planning are invisible. We present RHINO-AR, an interactive Augmented Reality (AR) museum exhibit that reintroduces the historical mobile robot RHINO into its original exhibition environment at the Deutsches Museum Bonn. To make the autonomous behavior that once made RHINO historically significant accessible again, our previous work, RHINO-VR, reconstructed the robot and its environment in a fully virtual museum scene. While engaging, this fully virtual presentation separated visitors from the real exhibition space and from the physical robot on display. RHINO-AR closes this reality gap by placing a virtual reconstruction of RHINO directly into the real museum space. Implemented on a Magic Leap 2 headset using Unity, our system combines real-time environment meshing with interactive visualizations of LiDAR sensing, traversability, and path planning to make otherwise invisible robotics processes understandable to non-expert visitors. We evaluated RHINO-AR in a two-day museum study with 22 participants, assessing usability, technical performance, satisfaction, conceptual understanding, and preference compared to RHINO-VR. The results indicate that RHINO-AR was well received, conveyed the targeted navigation concepts to a non-expert audience, and was generally preferred over the VR exhibit due to its stronger physical grounding and increased realism.

cs.RO↗