Summary

MemT-1.0-Nano is a memory foundation model designed to retain useful history as persistent model state. It combines dynamic local memory and trainable hyper memory so historical information can influence inference without repeatedly loading the original context.

Key Features

Persistent dynamic memory.
Trainable hyper memory.
Constant-size memory state.
Mixture-of-Storage design.
Qwen3.6-based backbone.
Transformers model distribution.
Memory-focused evaluation.
Experience-guided agentic coding.

About MemT-1.0-Nano


It gives agent developers a way to explore persistent experience and memory-aware coding with a bounded memory state. Core capabilities include Persistent dynamic memory, Trainable hyper memory, Constant-size memory state, and Mixture-of-Storage design.


Public code or model weights are linked from the project, with setup instructions for running or adapting the system. Additional characteristics include Qwen3.6-based backbone, Transformers model distribution, Memory-focused evaluation, and Experience-guided agentic coding.

Get more likes & reach the top of search results by adding this button on your site!

Embed button preview - Light theme
Embed button preview - Dark theme
TurboType Banner

Subscribe to the AI Search Newsletter

Get top updates in AI to your inbox every weekend. It's free!