mlsanityData to 14 Oct 2026

Action Deviation-Aware Inference for Low-Latency Wireless Robots

arXiv 2510.02851 ↗First posted 3 Oct 2025cs.RO cs.DCcited by 0

On the map: Efficient LLM inference and serving → Visual token pruning in multimodal models

Abstract (arXiv)

To support latency-sensitive AI applications ranging from autonomous driving to industrial robot manipulation, 6G envisions distributed ML with computational resources in mobile, edge, and cloud connected over hyper-reliable low-latency communication (HRLLC). In this setting, speculative decoding can facilitate collaborative inference of models distributively deployed: a lightweight on-device model locally generates drafts while a more capable remote target model on a server verifies and corrects them in parallel with speculative sampling, thus resulting in lower latency without compromising accuracy. However, unlike autoregressive text generation, behavior cloning policies, typically used for embodied AI applications, cannot parallelize verification and correction for multiple drafts as each generated action depends on observation updated by a previous action. To this end, we propose Action Deviation-Aware Hybrid Inference (ADAHI), wherein drafts are selectively transmitted and verified based on action deviation, which has a strong correlation with action's rejection probability by the target model. By invoking server operation only when necessary, communication and computational overhead can be reduced while accuracy gain from speculative sampling is preserved. Experiments on our testbed show that ADAHI reduces transmission and server operations by approximately 40%, lowers end-to-end latency by 39.2%, and attains up to 97.2% of the task-success rate of baseline that invokes speculative sampling for every draft embedding vector.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Visual token pruning in multimodal models, one of 12 directions in Efficient LLM inference and serving. Builds on 11 of its 19 references that are in the map. First posted on arXiv 3 Oct 2025.

Builds on

11 of 19 references.

References · 11

  1. 2022Fast Inference from Transformers via Speculative Decodingcited by 1,653
  2. 2021SoundStream: An End-to-End Neural Audio Codeccited by 771
  3. 2022Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulationcited by 551
  4. 2021Implicit Behavioral Cloningcited by 452
  5. 2024Behavior Generation with Latent Actionscited by 160
  6. 2024Mobile Edge Intelligence for Large Language Models: A Contemporary Surveycited by 128
  7. 2024Uncertainty-Aware Hybrid Inference with On-Device Small and Remote Large Language Modelscited by 19
  8. 2023Language-Oriented Communication with Semantic Coding and Knowledge Distillation for Text-to-Image Generationcited by 17
  9. 20256G Infrastructures for Edge AI: An Analytical Perspectivecited by 1
  10. 2025Energy-Efficient Wireless LLM Inference via Uncertainty and Importance-Aware Speculative Decodingcited by 1
  11. 2025Deadline-Aware Bandwidth Allocation for Semantic Generative Communication with Diffusion Modelscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026ComVLA: Communication-Aware Split Inference for VLA Models in 6G-Connected Robotics
  2. 2026Latent Action as Intention Enables Efficient Future Imagination for World Action Models
  3. 2026Algorithm-Architecture Co-Design for Efficient VLA Inference via Speculative Inference and Verification
  4. 2026World Action Models in Real Time: An Empirical Study of Smooth Execution via Asynchronous Deployment
  5. 2026AOSpec: Action and Observation Co-Speculation for Low-Latency Agent Serving
  6. 2026DREAM-Chunk: Reactive Action Chunking with Latent World Model
  7. 2026KERV: Kinematic-Rectified Speculative Decoding for Embodied VLA Models
  8. 2026Realtime-VLA FLASH: Speculative Inference Framework for Diffusion-based VLAs