From Inference Efficiency to Embodied Efficiency: Revisiting Efficiency Metrics for Vision-Language-Action Models

• URL: http://arxiv.org/abs/2603.19131v1
• Date: Thu, 19 Mar 2026 16:49:28 GMT
• Title: From Inference Efficiency to Embodied Efficiency: Revisiting Efficiency Metrics for Vision-Language-Action Models
• Abstract summary: Vision-Language-Action (VLA) models have recently enabled embodied agents to perform increasingly complex tasks.<n>We find that the prevailing notion of efficiency'' in current VLA research does not reflect actual performance on robotic platforms.
• Score: 5.744219633980964
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Vision-Language-Action (VLA) models have recently enabled embodied agents to perform increasingly complex tasks by jointly reasoning over visual, linguistic, and motor modalities. However, we find that the prevailing notion of ``efficiency'' in current VLA research, characterized by parameters, FLOPs, or token decoding throughput, does not reflect actual performance on robotic platforms. In real-world execution, efficiency is determined by system-level embodied behaviors such as task completion time, trajectory smoothness, cumulative joint rotation, and motion energy. Through controlled studies across model compression, token sparsification, and action sequence compression, we make several observations that challenge common assumptions. (1) Methods that reduce computation under conventional metrics often increase end-to-end execution cost or degrade motion quality, despite maintaining task success rates. (2) System-level embodied efficiency metrics reveal performance differences in the learned action policies that remain hidden under conventional evaluations. (3) Common adaptation methods such as in-context prompting or supervised fine-tuning show only mild and metric-specific improvements in embodied efficiency. While these methods can reduce targeted embodied-efficiency metrics such as jerk or action rate, the resulting gains may come with trade-offs in other metrics, such as longer completion time. Taken together, our results suggest that conventional inference efficiency metrics can overlook important aspects of embodied execution. Incorporating embodied efficiency provides a more complete view of policy behavior and practical performance, enabling fairer and more comprehensive comparisons of VLA models.

Related papers

• Invertible Neural Network Adapter for One-Step Flow Matching in Robot Manipulation [48.77928097573051]
  This paper presents an invertible neural network adapter for general robotic manipulation.<n>Built upon a flow-matching formulation, the proposed adapter constrains the action generation trajectory within an invertible latent space.<n>Experiments are conducted across a diverse set of simulation benchmarks and real-world robotic platforms to evaluate the effectiveness of the proposed method.
  arXiv  Detail & Related papers  (2026-06-17T15:35:27Z)
• Latent Action Reparameterization for Efficient Agent Inference [56.42014061367112]
  We propose a framework that learns a compact latent action space in which each latent action corresponds to a multi-step semantic behavior.<n>Unlike hand-crafted macros or hierarchical controllers, latent actions are learned from agent trajectories and integrated directly into the model.
  arXiv  Detail & Related papers  (2026-05-18T16:07:44Z)
• Act, Think or Abstain: Complexity-Aware Adaptive Inference for Vision-Language-Action Models [7.802379200026965]
  We propose an adaptive framework that dynamically routes VLA execution based on the complexity of the perceived state.<n>Our approach transforms the VLA's vision-language backbone into an active detection tool by projecting latent embeddings into an ensemble of parametric and non-parametric estimators.
  arXiv  Detail & Related papers  (2026-03-05T13:14:41Z)
• AC^2-VLA: Action-Context-Aware Adaptive Computation in Vision-Language-Action Models for Efficient Robotic Manipulation [21.23747444669735]
  We propose Action-Context-aware Adaptive Computation for VLA models (AC2-VLA)<n>AC2-VLA adaptively performs cognition reuse across timesteps, token pruning, and selective execution of model components within a unified mechanism.<n>Experiments on robotic manipulation benchmarks show that AC2-VLA achieves up to a 1.79times speedup while reducing FLOPs to 29.4% of the dense baseline.
  arXiv  Detail & Related papers  (2026-01-27T14:10:39Z)
• Toward Efficient Agents: Memory, Tool learning, and Planning [96.93533945696156]
  This paper investigates efficiency from three core components of agents: memory, tool learning, and planning, considering costs such as latency, tokens, steps, etc.
  arXiv  Detail & Related papers  (2026-01-20T17:51:56Z)
• PosA-VLA: Enhancing Action Generation via Pose-Conditioned Anchor Attention [92.85371254435074]
  PosA-VLA framework anchors visual attention via pose-conditioned supervision, consistently guiding the model's perception toward task-relevant regions.<n>We show that our method executes embodied tasks with precise and time-efficient behavior across diverse robotic manipulation benchmarks.
  arXiv  Detail & Related papers  (2025-12-03T12:14:29Z)
• EfficientFlow: Efficient Equivariant Flow Policy Learning for Embodied AI [8.90880897951322]
  Generative modeling has recently shown remarkable promise for visuomotor policy learning.<n>We introduce EfficientFlow, a unified framework for efficient embodied AI with flow-based policy learning.
  arXiv  Detail & Related papers  (2025-12-01T18:59:59Z)
• ActDistill: General Action-Guided Self-Derived Distillation for Efficient Vision-Language-Action Models [14.202025149504715]
  We present ActDistill, a framework that transfers the action prediction capability of any existing VLA model to a lightweight counterpart.<n>We employ a well-trained VLA model as the teacher and introduce a graph-structured encapsulation strategy to explicitly model the hierarchical evolution of action prediction.<n>Experiments on embodied benchmarks demonstrate that ActDistill achieves comparable or superior performance to full-scale VLA models while reducing computation by over 50% with up to 1.67 times speedup.
  arXiv  Detail & Related papers  (2025-11-22T14:44:03Z)
• Technical Debt in In-Context Learning: Diminishing Efficiency in Long Context [24.905102026459428]
  Transformers have demonstrated remarkable in-context learning capabilities, adapting to new tasks by simply conditioning on demonstrations without parameter updates.<n>It remains unclear to what extent the transformers optimally learn in-context compared to principled learning algorithms.
  arXiv  Detail & Related papers  (2025-02-07T00:26:45Z)
• ACE : Off-Policy Actor-Critic with Causality-Aware Entropy Regularization [52.5587113539404]
  We introduce a causality-aware entropy term that effectively identifies and prioritizes actions with high potential impacts for efficient exploration. Our proposed algorithm, ACE: Off-policy Actor-critic with Causality-aware Entropy regularization, demonstrates a substantial performance advantage across 29 diverse continuous control tasks.
  arXiv  Detail & Related papers  (2024-02-22T13:22:06Z)
• Building an Efficiency Pipeline: Commutativity and Cumulativeness of Efficiency Operators for Transformers [68.55472265775514]
  We consider an efficiency method as an operator applied on a model. In this paper, we study the plausibility of this idea, and the commutativity and cumulativeness of efficiency operators.
  arXiv  Detail & Related papers  (2022-07-31T18:01:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.