Newest papers

MemT-1.0: Towards Memory-Native Evolution

Zhiyu Li, Zeyu Zhang, Linyun Liu, Yiwei Bu, Chenyang Xi, Xichong Zhang, Yi Yu, Yu Yu, Zhichen Liu, Yuxiao Yi, Chunyu Li, Haoran Liu, Zhenkai Ding, Le Zhou, Juncheng Zhang, Boyu Chen, Zehao Lin, Guorui Li, Renyu Fu, Ji Li · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.26825 · cited by 0 works (OpenAlex) · open-access PDF

French

À cette fin, nous présentons MemT-1.0, une nouvelle architecture de modèles de fondations de mémoire qui internalisent la mémoire dans le cadre du calcul des modèles. Construit sur Metis, MemT-1.0 emploie Gated DeltaNet-2 avec un mécanisme de mise à jour de mémoire stable au niveau des jetons au sein de chaque tour. Nous introduisons également Mixture-of-Storage pour évoluer la capacité de mémoire native. Après une préparation délibérée de la base, nous développons un pipeline de formation en deux étapes consistant en un pré-entraînement continu et un post-entraînement, soutenu par 34B de jetons de tâches de mémoire de formation et une infrastructure de formation et d'inférence efficace. Nous lançons un modèle de famille compren

Machine translation (Meta M2M100); the English original follows.

Abstract

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 $1.86\times$ end-to-end inference speedup and a $66\times$ 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.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Computer science, Inference, Task (project management), Pipeline (software), Context (archaeology), Memory model

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange