← sources

[Paper] MemT-1.0: Towards Memory-Native Evolution

source: reddit (reddit:LocalLLaMA)published 2026-10-21 12:00ingested 2026-10-22 02:43

retained text

>Memory-native evolution requires making memory an intrinsic capability of foundation models, enabling them to continuously accumulate and utilize information. To this end, we introduce MemT-1.0, a new architecture of memory foundation models that internalize memory as part of model computation. Building on Metis, MemT-1.0 employs Gated DeltaNet-2 with a stable token-level memory update mechanism within each turn. We further introduce Mixture-of-Storage to scale native memory capacity. Following a deliberate backbone preparation, we develop a two-stage training pipeline consisting of continued pre-training and post-training, supported by 34B tokens of training data and efficient training and inference infrastructure. We release a model family comprising MemT-1.0-Nano and MemT-1.0-Flash, and conduct a comprehensive evaluation of memory and general capabilities. The results show that they outperform the parametric memory baselines on average and remain competitive with retrieval-augmented methods, while largely preserving general capabilities. **At a history length of 128K tokens, MemT-1.0-Nano achieves approximately 1.86× end-to-end inference speedup and a 66× reduction in persistent storage relative to full context KV cache**. By internalizing retrieved procedural experience, MemT-1.0 improves task success on agentic tasks, providing a practical step toward memory-native evolution.

* **arXiv** : [https://arxiv.org/abs/2610.26825](https://arxiv.org/abs/2610.26825)
* **PDF** : [https://arxiv.org/pdf/2610.26825](https://arxiv.org/pdf/2610.26825)
* **GitHub** : [https://github.com/MemTensor/MemT-1.0](https://github.com/MemTensor/MemT-1.0)
* **HuggingFace** : [https://huggingface.co/collections/MemTensor/memt-1](https://huggingface.co/collections/MemTensor/memt-1)

document fetched 2026-10-22 00:45 · canonical https://reddit.com/r/LocalLLaMA/comments/1xbl751/paper_memt10_towards_memorynative_evolution