[2604.01577] Thinking While Listening: Fast-Slow Recurrence for Long-Horizon Sequential Modeling

Computer Science > Machine Learning arXiv:2604.01577 (cs) [Submitted on 2 Apr 2026] Title:Thinking While Listening: Fast-Slow Recurrence for Long-Horizon Sequential Modeling Authors:Shota Takashiro, Masanori Koyama, Takeru Miyato, Yusuke Iwasawa, Yutaka Matsuo, Kohei Hayashi View a PDF of the paper titled Thinking While Listening: Fast-Slow Recurrence for Long-Horizon Sequential Modeling, by Shota Takashiro and 5 other authors View PDF HTML (experimental) Abstract:We extend the recent latent recurrent modeling to sequential input streams. By interleaving fast, recurrent latent updates with self-organizational ability between slow observation updates, our method facilitates the learning of stable internal structures that evolve alongside the input. This mechanism allows the model to maintain coherent and clustered representations over long horizons, improving out-of-distribution generalization in reinforcement learning and algorithmic tasks compared to sequential baselines such as LSTM, state space models, and Transformer variants. Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2604.01577 [cs.LG]   (or arXiv:2604.01577v1 [cs.LG] for this version)   https://doi.org/10.48550/arXiv.2604.01577 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Shota Takashiro [view email] [v1] Thu, 2 Apr 2026 03:45:13 UTC (1,160 KB) Full-text links: Access Paper: View a PDF of the paper titled Thinking While Listening: Fast-Slow Recurr...