mlsanityData to 7 Oct 2026

D2-VLA: Dual-Memory Dual-Frequency Vision-Language-Action Model For Long Dynamic Manipulation

arXiv 2609.34792 ↗First posted 28 Sep 2026cs.CVcited by 0

On the map: Vision-language-action models → Reasoning and chain-of-thought in VLA also fits: Efficient VLA inference and adaptation

Abstract (arXiv)

Long-horizon manipulation requires robots to remember cues that are no longer in view while responding to moving objects. Yet vision-language-action (VLA) policies often rely on the latest observation, and refreshing their visual context typically requires another costly vision-language model (VLM) pass. We present D2-VLA, which combines dual memory and dual-frequency control at the KV-cache interface of a pretrained VLA. D2-VLA uses block-wise causal KV caching to encode observations incrementally and, guided by distinct temporal attention patterns, constructs separate historical KV read views for the VLM and action expert. Between periodic VLM updates, a gated adapter incorporates fresh visual features into the latest history-conditioned KV block, while a short fast-memory queue supports action replanning. We introduce DOMINO-Long, a ten-task benchmark requiring robots to use earlier visual cues when manipulating moving objects. D2-VLA achieves complete-task success rates of 29.3% on DOMINO, compared with 9.6% for π0.5 and 17.2% for PUMA, and 60.0% on DOMINO-Long, compared with 35.4% and 20.6%, respectively. It improves success rates on eight real-robot tasks and reaches 97.5% on LIBERO-Long and 74.3% on RoboTwin 2.0.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reasoning and chain-of-thought in VLA, one of 13 directions in Vision-language-action models. Builds on 29 of its 33 references that are in the map. First posted on arXiv 28 Sep 2026.

Builds on

29 of 33 references.

Closest ancestors · 4

Same direction, newest first

  1. 2026ReMem-VLA: Empowering Vision-Language-Action Model with Memory via Dual-Level Recurrent Queries13 Mar
  2. 2026TempoFit: Plug-and-Play Layer-Wise Temporal KV Memory for Long-Horizon Vision-Language-Action Manipulation8 Mar
  3. 2026RoboMME: Benchmarking and Understanding Memory for Robotic Generalist Policies4 Mar
  4. 2025CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Models27 Mar

Other references · 25

  1. 2023Sigmoid Loss for Language Image Pre-Trainingcited by 3,048
  2. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
  3. 2025π0.5: a Vision-Language-Action Model with Open-World Generalizationcited by 1,653
  4. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
  5. 2025Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Successcited by 844
  6. 2024PaliGemma: A versatile 3B VLM for transfercited by 679
  7. 2024RDT-1B: a Diffusion Foundation Model for Bimanual Manipulationcited by 670
  8. 2025FAST: Efficient Action Tokenization for Vision-Language-Action Modelscited by 643
  9. 2020SAPIEN: A SimulAted Part-based Interactive ENvironmentcited by 584
  10. 2025SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Roboticscited by 507
  11. 2025RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulationcited by 477
  12. 2023Vision-Language Foundation Models as Effective Robot Imitatorscited by 344
13 more
  1. 2025Real-Time Execution of Action Chunking Flow Policiescited by 199
  2. 2025VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Modelcited by 140
  3. 2024FlowPolicy: Enabling Fast and Robust 3D Flow-based Policy via Consistency Flow Matching for Robot Manipulationcited by 99
  4. 2024HiRT: Enhancing Robotic Control with Hierarchical Robot Transformerscited by 72
  5. 2025MemER: Scaling Up Memory for Robot Control via Experience Retrievalcited by 64
  6. 2026AHA-WAM:Asynchronous Horizon-Adaptive World-Action Modeling with Observation-Guided Context Routingcited by 41
  7. 2026DynamicVLA: A Vision-Language-Action Model for Dynamic Object Manipulationcited by 33
  8. 2026StarVLA-α: Reducing Complexity in Vision-Language-Action Systemscited by 27
  9. 2025CronusVLA: Towards Efficient and Robust Manipulation via Multi-Frame Vision-Language-Action Modelingcited by 24
  10. 2026Towards Generalizable Robotic Manipulation in Dynamic Environmentscited by 17
  11. 2026FAVLA: A Force-Adaptive Fast-Slow VLA model for Contact-Rich Robotic Manipulationcited by 14
  12. 2026TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with <1 GB VRAMcited by 9
  13. 2026SeedPolicy: Horizon Scaling via Self-Evolving Diffusion Policy for Robot Manipulationcited by 4

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025VLA-Cache: Efficient Vision-Language-Action Manipulation via Adaptive Token Caching
  2. 2026ChainVLA: Chaining Vision-Language-Action Queries through a Unified Execution State for Long-Horizon Manipulation
  3. 2025Long-VLA: Unleashing Long-Horizon Capability of Vision Language Action Model for Robot Manipulation
  4. 2026AC^2-VLA: Action-Context-Aware Adaptive Computation in Vision-Language-Action Models for Efficient Robotic Manipulation
  5. 2026S2-VLA: State-Space Guided Vision-Language-Action Models for Long-Horizon Manipulation
  6. 2026Explicit Language Memory for Long-Horizon Planning in Vision-Language-Action Models
  7. 2026EvoMem-VLA: State-Evolution Memory for Long-Horizon Robot Manipulation
  8. 2026Reducing Temporal Redundancy for Efficient Vision-Language-Action Inference