Atsushi Saito


2023

pdf
RWKV: Reinventing RNNs for the Transformer Era
Bo Peng | Eric Alcaide | Quentin Anthony | Alon Albalak | Samuel Arcadinho | Stella Biderman | Huanqi Cao | Xin Cheng | Michael Chung | Leon Derczynski | Xingjian Du | Matteo Grella | Kranthi Gv | Xuzheng He | Haowen Hou | Przemyslaw Kazienko | Jan Kocon | Jiaming Kong | Bartłomiej Koptyra | Hayden Lau | Jiaju Lin | Krishna Sri Ipsit Mantri | Ferdinand Mom | Atsushi Saito | Guangyu Song | Xiangru Tang | Johan Wind | Stanisław Woźniak | Zhenyuan Zhang | Qinghua Zhou | Jian Zhu | Rui-Jie Zhu
Findings of the Association for Computational Linguistics: EMNLP 2023

Transformers have revolutionized almost all natural language processing (NLP) tasks but suffer from memory and computational complexity that scales quadratically with sequence length. In contrast, recurrent neural networks (RNNs) exhibit linear scaling in memory and computational requirements but struggle to match the same performance as Transformers due to limitations in parallelization and scalability. We propose a novel model architecture, Receptance Weighted Key Value (RWKV), that combines the efficient parallelizable training of transformers with the efficient inference of RNNs. Our approach leverages a linear attention mechanism and allows us to formulate the model as either a Transformer or an RNN, thus parallelizing computations during training and maintains constant computational and memory complexity during inference. We scale our models as large as 14 billion parameters, by far the largest dense RNN ever trained, and find RWKV performs on par with similarly sized Transformers, suggesting future work can leverage this architecture to create more efficient models. This work presents a significant step towards reconciling trade-offs between computational efficiency and model performance in sequence processing tasks.

2018

pdf
Curriculum Learning Based on Reward Sparseness for Deep Reinforcement Learning of Task Completion Dialogue Management
Atsushi Saito
Proceedings of the 2018 EMNLP Workshop SCAI: The 2nd International Workshop on Search-Oriented Conversational AI

Learning from sparse and delayed reward is a central issue in reinforcement learning. In this paper, to tackle reward sparseness problem of task oriented dialogue management, we propose a curriculum based approach on the number of slots of user goals. This curriculum makes it possible to learn dialogue management for sets of user goals with large number of slots. We also propose a dialogue policy based on progressive neural networks whose modules with parameters are appended with previous parameters fixed as the curriculum proceeds, and this policy improves performances over the one with single set of parameters.