Axiom-0: Hierarchical Self-Refinement for Vision-Language-Action Policies

Renbin Shu; Jing Yang; Xiangyuan You; Guolin Qiu; Yufei Xia; Yiming Zhu; Nanjun Dai; Fei Zhang

cs.ROOct 9, 2026

Axiom-0: Hierarchical Self-Refinement for Vision-Language-Action Policies

Authors: Renbin Shu, Jing Yang, Xiangyuan You, Guolin Qiu, Yufei Xia, Yiming Zhu, Nanjun Dai, Fei Zhang, +2 more

Organizations: 4Axiom Robotics Team

Abstract

Reliable manipulation requires policies that generalize across tasks, layouts, and object interactions. Vision-language-action (VLA) models combine pretrained multimodal understanding with learned control, but their performance remains limited on unfamiliar task compositions. We present Axiom-0, a VLA policy that addresses both the quality of its representations and their use in action generation through hierarchical self-refinement. Intra-VLM Representation Refinement (IVRR) strengthens selected intermediate features through consistency supervision with learned feature projections during training. VLM-Action Interface Refinement (VAIR) recursively updates the resulting conditioning before action generation. Joint optimization of the supervised VLM layers, VAIR, and the action generator connects the two components. On RoboCasa365, Axiom-0 achieves 62.32% overall success in our evaluation, exceeding the highest published leaderboard score (Paimon-0, 58.1%) by 4.22 percentage points and the Xiaomi-Robotics-1 reference by 4.92 points. Under the evaluated profiling setup, mean action-chunk generation latency increases by 1.7% over Xiaomi-Robotics-1. These results demonstrate improved task completion with modest inference overhead.

Figures & tables

Figure 1: Overview of hierarchical self-refinement in Axiom-0. Left: IVRR aligns projected intermediate VLM features with spatial foundation-model references during training. Right: VAIR recursively refines the conditioning supplied to the action generator using a shared refinement module. Common inputs, the VLM backbone, and the DiT action generator lie outside the refinement panels. Solid arrows denote inference paths; dashed arrows denote auxiliary training paths. Feature grids are schematic. Figure 2 expands the VAIR computation and its integration with action sampling.

Figure 2: VAIR computation and integration with the action generator. (a)–(c) show the added refinement computation: values are projected into a compact latent space, updated by a shared Transformer over

R

iterations, and converted into a gated residual correction to the original values. The subtraction in (c) takes

Z_{l}^{(R)}

as its input and

Z_{l}^{(0)}

as its reference. Keys pass through unchanged. (d) shows the refined cache and the DiT sampling path; the cache is computed once per observation and reused across five Euler sampling steps. Feature grids and Transformer internals are schematic.

Method	Atomic Seen	Composite Seen	Composite Unseen	Overall
Paimon-0	79.3	55.8	36.7	58.1
Xiaomi-Robotics-1	80.2	57.1	32.1	57.4
Phasor-m7	78.7	57.4	31.5	56.8
ProWAM	76.0	46.5	18.2	48.1
PatchWAM	78.7	46.4	13.8	47.6
ABot-M0.6	79.4	48.3	7.9	46.6

Table 1: RoboCasa365 benchmark comparison. Success rates (%), using the published leaderboard snapshot of October 3, 2026 [ 12 ] . Ours is author-evaluated. Bold marks the largest reported point estimate in each column.

IVRR	VAIR	Atomic Seen	Composite Seen	Composite Unseen	Overall	$\Delta$
–	–	80.20	57.10	32.10	57.40	–
$\checkmark$	–	84.11	56.38	37.50	60.32	+2.92
–	$\checkmark$	83.78	60.125	35.50	60.76	+3.36
$\checkmark$	$\checkmark$	86.22	60.12	37.62	62.32	+4.92

Table 2: Component configurations on RoboCasa365. IVRR: Intra-VLM Representation Refinement; VAIR: VLM–Action Interface Refinement. Success rates are reported in %;

\Delta

denotes the overall gain in percentage points over the published Xiaomi-Robotics-1 reference [ 15 ] . Split scores retain their reported precision. Comparisons reflect the resulting trained configurations.

Metric	Xiaomi-Robotics-1	Ours	Change
Mean latency (ms)	133.25	135.55	+2.30 (1.73%)
Median latency (ms)	131.76	134.78	+3.02 (2.29%)
P95 latency (ms)	139.19	140.53	+1.34 (0.96%)
Parameters (B)	5.11322	5.12791	+14.684 M (0.287%)
Estimated compute (TFLOPs/chunk)	4.751	4.779	+0.028 (0.59%)
Peak GPU memory (GiB)	9.709	9.763	+0.054 (0.56%)

Table 3: Inference cost on an NVIDIA RTX PRO 5000 (72 GB). Batch size is one; each call produces a 16-action chunk. B and M denote billions and millions of parameters.

Explore similar work

AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Models
Mar 10, 2026Yutong Hu, Jan-Nico Zaech, Nikolay Nikolov +6Vision-Language-Action Models Autoregressive Generation
∗INSAIT, Sofia University “St. Kliment Ohridski” · †KU Leuven, Dept. Mechanical Engineering, Research unit Robotics, Automation and Mechatronics · ‡KU Leuven, Dept. Electrical Engineering, Research unit Processing Speech and Images
Not All Features Are Created Equal: A Mechanistic Study of Vision-Language-Action Models
Mar 19, 2026Bryce Grant, Xijia Zhao, Peng WangVision-Language-Action Models Robotics
Case Western Reserve University
TokenPB: Enabling VLA Action Self-Verification via VLM Token Probability Bucketing
Mar 18, 2026Chen Zhao, Zhuoran Wang, Haoyang Li +6Vision-Language-Action Models
School of Information, Renmin University of China · Key Laboratory of Data Engineering and Knowledge Engineering, Beijing, China · Beijing University of Posts and Telecommunications +1

Mar 10, 2026J/K move · Enter open · S save

AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Models

Yutong Hu, Jan-Nico Zaech, Nikolay Nikolov, Yuanqi Yao+5

∗INSAIT, Sofia University “St. Kliment Ohridski” · †KU Leuven, Dept. Mechanical Engineering, Research unit Robotics, Automation and Mechatronics · ‡KU Leuven, Dept. Electrical Engineering, Research unit Processing Speech and Images

Vision-Language-Action Models Autoregressive Generation Robotic Manipulation

Open paper AskPDF arXiv

We propose a standalone autoregressive (AR) Action Expert that generates actions as a continuous causal sequence while conditioning on refreshable vision-language prefixes. In contrast to existing Vision-Language-Action (VLA) models and diffusion policies that reset temporal context with each new observation and predict actions reactively, our Action Expert maintains its own history through a long-lived memory and is inherently context-aware. This structure addresses the frequency mismatch between fast control and slow reasoning, enabling efficient independent pretraining of kinematic syntax and modular integration with heavy perception backbones, naturally ensuring spatio-temporally consistent action generation across frames. To synchronize these asynchronous hybrid V-L-A modalities, we utilize a re-anchoring mechanism that mathematically accounts for perception staleness during both training and inference. Experiments on simulated and real-robot manipulation tasks demonstrate that the proposed method can effectively replace traditional chunk-based action heads for both specialist and generalist policies. AR-VLA exhibits superior history awareness and substantially smoother action trajectories while maintaining or exceeding the task success rates of state-of-the-art reactive VLAs. Overall, our work introduces a scalable, context-aware action generation schema that provides a robust structural foundation for training effective robotic policies. Code and Videos available at https://arvla.insait.ai

Abstract

Explore similar work

AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Models

Not All Features Are Created Equal: A Mechanistic Study of Vision-Language-Action Models

TokenPB: Enabling VLA Action Self-Verification via VLM Token Probability Bucketing