๐
๐
The Cartographer
Bilinear Input Modulation for Mamba: Koopman Bilinear Forms for Memory Retention and Multiplicative Computation
April 19, 2026 ยท Grace Period ยท + Add venue
Authors
Hiroki Fujii, Masaki Yamakita
arXiv ID
2604.17221
Category
eess.SY: Systems & Control (EE)
Cross-listed
cs.LG,
math.DS
Citations
0
Abstract
Selective State Space Models (SSMs), notably Mamba, employ diagonal state transitions that limit both memory retention and bilinear computational capacity. We propose a factorized bilinear input modulation that augments the SSM with a state-input product, interpretable as a finite-dimensional Koopman bilinear form. After introducing a shared state across channels (Coupled SSM), the modulation admits two implementations. Coupled Bilinear Input Modulation (Coupled-BIM) retains the full bilinear product at the cost of sequential computation, while Coupled Gated Modulation (Coupled-GM) linearizes it into a gate modulation that is compatible with the parallel scan. Experiments on a multiple input-delay pendulum (memory retention) and NARMA-10 (bilinear computation) reveal a clear dissociation. Coupled-GM substantially improves memory retention but not bilinear computation, while Coupled-BIM improves both. A pathway ablation confirms that the two downstream routes of the bilinear signal serve complementary roles. The improvement is statistically robust, with Coupled-BIM consistently outperforming all other variants on bilinear computation. Furthermore, only Coupled-BIM benefits from increasing the SSM state dimension, while coupling or gate modulation alone show no improvement, establishing the bilin-ear mechanism as uniquely capable of exploiting larger state spaces.
Community Contributions
Found the code? Know the venue? Think something is wrong? Let us know!
๐ Similar Papers
In the same crypt โ Systems & Control (EE)
๐
๐
The Cartographer
Incremental Gradient, Subgradient, and Proximal Methods for Convex Optimization: A Survey
๐
๐
The Cartographer
Wireless Network Design for Control Systems: A Survey
R.I.P.
๐ป
Ghosted
Learning-based Model Predictive Control for Safe Exploration
R.I.P.
๐ป
Ghosted
Safety-Critical Model Predictive Control with Discrete-Time Control Barrier Function
R.I.P.
๐ป
Ghosted