yifanzhang-pro/recurrent-looped-tranformer

Implementation of the recurrent looped transformer proposed by Yifan Zhang of Princeton

Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computation (NeurIPS 2025)

Official Project Page for Deep Delta Learning (https://arxiv.org/abs/2601.00417)

Official repository of the xLSTM.

RWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RNN and transformer - great performance, linear time, constant space (no kv-cache), fast training, infinite ctx_len, and free sentence embedding.

FlashLoop: Fast and Memory-Efficient Looped Transformers via Lazy Updates

A concise but complete full-attention transformer with a set of promising experimental features from various papers