[2603.22333] Graph Signal Processing Meets Mamba2: Adaptive Filter Bank via Delta Modulation

Computer Science > Machine Learning arXiv:2603.22333 (cs) [Submitted on 20 Mar 2026] Title:Graph Signal Processing Meets Mamba2: Adaptive Filter Bank via Delta Modulation Authors:Yehjin Shin, Seojin Kim, Noseong Park View a PDF of the paper titled Graph Signal Processing Meets Mamba2: Adaptive Filter Bank via Delta Modulation, by Yehjin Shin and 2 other authors View PDF HTML (experimental) Abstract:State-space models (SSMs) offer efficient alternatives to attention with linear-time recurrence. Mamba2, a recent SSM-based language model, uses selective input gating and a multi-head structure, enabling parallel computation and strong benchmark performance. However, its multi-head recurrence operates independently without structured utilization or analysis. In this work, we propose a novel method called Hierarchical ADaptive filter bank for Efficient SSMs (HADES), a Graph Signal Processing (GSP)-inspired framework that reinterprets Mamba2 as an adaptive filter bank on a line graph. Our hierarchical architecture introduces two filter types: shared filters for global low-pass behavior and expert filters for local high-pass behavior, achieved through structured bias on the parameter {\Delta}. HADES achieves comparable performance to baseline models including Mamba2 across various benchmarks in language modeling, commonsense reasoning, and long-context retrieval, while using only 58.9% of the original parameters. In this regard, HADES bridges GSP and neural sequence modeling, enab...