The official implementation of the paper "StateX: Enhancing RNN Recall via Post-training State Expansion". - View it on GitHub