Unofficial Implementation of Titans: Learning to Memorize at Test Time

a JAX + NNX implementation of Google's Titan NN. Transformer short-term attention + neural long-term memory with test-time updates

This repository contains an experimental implementation of the Titans Transformer architecture for sequence modeling tasks. The code is a personal exploration and may include errors or inefficiencies as I am currently in the learning stage. It is inspired by the ideas presented in the original

Complete PyTorch reproduction of Google's TITANS, MIRAS, and NL neural memory papers. 52 tests, 87% coverage, Docker support.

Visual animated walkthroughs of the DeepMind "Titans: Learning to Memorize at Test Time" paper using Manim, aimed at making complex ML concepts accessible.

Titans: Learning to Memorize at Test Time

Educational demo of Google's Titans surprise-based memory mechanism with PyTorch, interactive notebooks, and visualizations

Titans Match Stats Tracker

Titans sports stat trackers — Basketball & Soccer across all age groups

High-performance CUDA implementation of Titans neural memory architecture (Learning to Memorize at Test Time)

A CPU-only measurement instrument for the timescale spectrum and aging dynamics of a model's memory (continual learning / long-term memory). Is your memory really multi-timescale, or secretly single-speed? MIT.

A practical, educational repository demonstrating how to implement Test-Time Memorization using the Titans architecture with MIRAS training strategy.

Gerador de cards com interações, gerados pelo js.

NIFTY 100 equity forecasting with Google's Titans Memory- Augmented Transformer

PWA for real-time soccer match statistics tracking — Football Titans Tracker