NativeMEM: Native Memory Compression for Long-Horizon Robotic Manipulation

• URL: http://arxiv.org/abs/2607.06678v1
• Date: Tue, 07 Jul 2026 18:00:53 GMT
• Title: NativeMEM: Native Memory Compression for Long-Horizon Robotic Manipulation
• Abstract summary: We present NativeMEM, a policy that features long-term and real-time updated memory.<n>NativeMEM consistently outperforms prior methods, boosting success rates from 32.4% to 84.0% in simulation and up to 98.7% on real robots.
• Score: 22.20970271577043
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: How can pretrained Vision-Language-Action (VLA) models retain long-horizon visual histories with high-frequency updates without sacrificing efficiency? Existing approaches rely on external memory management, which restrains either the memory horizon or the reactiveness of pretrained policies. To this end, we present NativeMEM, a VLA policy that features long-term and real-time updated memory. At its core is an efficient memory encoding scheme, Native Memory Compression, which repurposes the VLA's own vision encoder to compress each historical frame from each camera view into a single token. Appended to the input sequence, these memory tokens enable the pretrained VLA to attend over long-term history with negligible latency overhead, requiring neither an external planner nor a freshly initialized memory module. To align the memory tokens with the pretrained policy, we first develop a generic memory tokenizer under the supervision of a frozen VLA on memory-demanding data, and then unfreeze the VLA for task-specific fine-tuning. NativeMEM consistently outperforms prior methods, boosting success rates from 32.4% to 84.0% in simulation and up to 98.7% on real robots, while maintaining low inference latency and GPU memory usage. Notably, NativeMEM exhibits high data efficiency by achieving competitive results with prior arts using only 20% of the training data.

Related papers

• UniMem: Unifying Multimodal Memory and Control for Vision-Language-Action Models [8.70663346892131]
  We present UniMem, a framework that unifies high-level, multimodal memory and low-level control under one backbone.<n>We evaluate UniMem across five simulation and four hardware tasks targeting sequential and spatial memory.
  arXiv  Detail & Related papers  (2026-08-24T06:56:02Z)
• Remember Smarter: Visual History Compressor and Hyperbolic Experience Space for Robotic Memory [54.400138177707504]
  Remember Smarter (RS) is a plug-and-play module with complementary visual-history and hyperbolic experience-memory branches.<n>When adapted to pi0, RS increases total success on LIBERO-Plus from 53.6% to 70.6% and achieves substantial performance gains in real-robot experiments.
  arXiv  Detail & Related papers  (2026-08-15T15:13:33Z)
• MemoryWAM: Efficient World Action Modeling with Persistent Memory [80.90899269062128]
  We introduce MemoryWAM, a world action model with efficient persistent memory.<n>A tailored attention mechanism enables retrieval of both detailed short-term context and compressed long-term context.<n>MemoryWAM outperforms strong vision-language-action (VLA) and WAM baselines in both simulation and the real world.
  arXiv  Detail & Related papers  (2026-06-18T17:59:51Z)
• FlashMemory-DeepSeek-V4: Lightning Index Ultra-Long Context via Lookahead Sparse Attention [77.12062766962815]
  Lookahead Sparse Attention (LSA) is powered by a Neural Memory Indexer built upon the DeepSeek-V4 architecture.<n>We instantiate this architecture via a backbone-free decoupled training strategy.<n>We show that FM-DS-V4 compresses the average physical KV cache footprint down to merely 13.5% of the full-context baseline.
  arXiv  Detail & Related papers  (2026-06-08T06:25:54Z)
• RoboMemArena: A Comprehensive and Challenging Robotic Memory Benchmark [49.16825786282095]
  Memory is a critical component of robotic intelligence, as robots must rely on past observations and actions to accomplish long-horizon tasks.<n>Existing robotic memory benchmarks still lack multimodal annotations for memory formation, provide limited task coverage and structural complexity, and remain restricted to simulation without real-world evaluation.<n>We address this gap with RoboMemArena, a large-scale benchmark of 26 tasks, with average trajectory lengths exceeding 1,000 steps per task and 68.9% of subtasks being memory-dependent.
  arXiv  Detail & Related papers  (2026-05-11T17:54:49Z)
• MSA: Memory Sparse Attention for Efficient End-to-End Memory Model Scaling to 100M Tokens [38.762431789628074]
  We present Memory Sparse Attention (MSA), an end-to-end trainable, efficient, and massively scalable memory model framework.<n>MSA achieves linear complexity in both training and inference while maintaining exceptional stability.<n>We also propose Memory Interleaving to facilitate complex multi-hop reasoning across scattered memory segments.
  arXiv  Detail & Related papers  (2026-03-06T02:29:54Z)
• VPWEM: Non-Markovian Visuomotor Policy with Working and Episodic Memory [31.464584758455356]
  VPWEM is a non-Markovian visuomotor policy equipped with working and episodic memories.<n>It exploits both short-term and episode-wide information for action generation with nearly constant memory and computation per step.
  arXiv  Detail & Related papers  (2026-03-05T07:52:50Z)
• LightMem: Lightweight and Efficient Memory-Augmented Generation [72.21680105265824]
  We introduce a new memory system called LightMem, which strikes a balance between the performance and efficiency of memory systems.<n>Inspired by the Atkinson-Shiffrin model of human memory, LightMem organizes memory into three complementary stages.<n>Experiments on LongMemEval with GPT and Qwen backbones show that LightMem outperforms strong baselines in accuracy (up to 10.9% gains) while reducing token usage by up to 117x.
  arXiv  Detail & Related papers  (2025-10-21T17:58:17Z)
• MemoryVLA: Perceptual-Cognitive Memory in Vision-Language-Action Models for Robotic Manipulation [59.31354761628506]
  Temporal context is essential for robotic manipulation because such tasks are inherently non-Markovian, yet mainstream VLA models typically overlook it.<n>We propose MemoryVLA, a Cognition-Memory-Action framework for long-horizon robotic manipulation.<n>We evaluate it on 150+ simulation and real-world tasks across three robots.
  arXiv  Detail & Related papers  (2025-08-26T17:57:16Z)
• MEMO: Fine-grained Tensor Management For Ultra-long Context LLM Training [24.066283519769968]
  Large Language Models (LLMs) have been trained using extended context lengths to foster more creative applications.<n>We propose MEMO, a novel framework for fine-grained activation memory management.<n>MeMO achieves an average of 1.97x and 1.80x MFU compared to Megatron-LM and DeepSpeed.
  arXiv  Detail & Related papers  (2024-07-16T18:59:49Z)

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.