Metraipolis › archive › dc_ce4e05c6742f
Variational Cognitive Memory Abstraction for Long-Horizon Vision-Language-Action Models — arXiv:2610.34251
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.34251
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-27T07:57:29.712Z
- SHA-256 of the archived text
b85862aa7b2aac4962444cce7f4f7d6f0027cc3beee1e4a6587de671dd42ee06- Stable URL of this page
https://metraipolis.com/documento/dc_ce4e05c6742f
How to cite
Variational Cognitive Memory Abstraction for Long-Horizon Vision-Language-Action Models — arXiv:2610.34251 — arXiv (en), https://arxiv.org/abs/2610.34251 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ce4e05c6742f (Metraipolis archive dc_ce4e05c6742f, sha256 b85862aa7b2aac4962444cce7f4f7d6f0027cc3beee1e4a6587de671dd42ee06)
BibTeX
@misc{dcce4e05c6742f,
title = {Variational Cognitive Memory Abstraction for Long-Horizon Vision-Language-Action Models — arXiv:2610.34251},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.34251},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ce4e05c6742f; sha256 b85862aa7b2aac4962444cce7f4f7d6f0027cc3beee1e4a6587de671dd42ee06},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.34251. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Vision-Language-Action (VLA) models have shown remarkable success in robotic manipulation, yet their ability to execute long-horizon tasks remains limited by insufficient long- term memory modeling. Existing methods either rely on implicit temporal representations within policy networks, where historical information can be gradually diluted, or maintain explicit memory buffers that suffer from redundancy, noise accumulation, and inefficient retrieval. In this paper, we propose CogMemory, a novel variational cognitive memory framework that enables VLA systems to selectively preserve and reuse task-relevant historical information. Our approach encodes multimodal trajectories into compact latent memory representations through a selective state- space memory encoder, multi-slot temporal abstraction, and variational information compression. During action generation, dynamic cross-attention retrieval with gated fusion adaptively incorporates relevant memories into policy prediction. With merely 3.45M additional parameters, CogMemory stays lightweight and seamlessly fits both early- and intermediate-fusion VLA backbones. Extensive experiments on long-horizon robotic manipulation benchmarks demonstrate consistent improvements across diverse VLA backbones. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.34251 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-27T07:55:30.683Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).