Knob: A Physics-Inspired Gating Interface for Interpretable and Controllable Neural Dynamics

TL;DR

Knob introduces a control-theoretic, physics-inspired gating mechanism for neural networks, enabling dynamic, interpretable, and controllable model behavior suitable for static and streaming tasks.

Contribution

It proposes a novel neural gating framework based on second-order control dynamics, linking physical parameters to neural calibration and enabling dual-mode inference.

Findings

Gate responses exhibit second-order control signatures

Framework allows intuitive tuning of stability and sensitivity

Validation on CIFAR-10-C confirms calibration and control properties

Abstract

Existing neural network calibration methods often treat calibration as a static, post-hoc optimization task. However, this neglects the dynamic and temporal nature of real-world inference. Moreover, existing methods do not provide an intuitive interface enabling human operators to dynamically adjust model behavior under shifting conditions. In this work, we propose Knob, a framework that connects deep learning with classical control theory by mapping neural gating dynamics to a second-order mechanical system. By establishing correspondences between physical parameters -- damping ratio ($\zeta$) and natural frequency ($\omega_n$) -- and neural gating, we create a tunable "safety valve". The core mechanism employs a logit-level convex fusion, functioning as an input-adaptive temperature scaling. It tends to reduce model confidence particularly when model branches produce conflicting predictions. Furthermore, by imposing second-order dynamics (Knob-ODE), we enable a \textit{dual-mode} inference: standard i.i.d. processing for static tasks, and state-preserving processing for continuous streams. Our framework allows operators to tune "stability" and "sensitivity" through familiar physical analogues. This paper presents an exploratory architectural interface; we focus on demonstrating the concept and validating its control-theoretic properties rather than claiming state-of-the-art calibration performance. Experiments on CIFAR-10-C validate the calibration mechanism and demonstrate that, in Continuous Mode, the gate responses are consistent with standard second-order control signatures (step settling and low-pass attenuation), paving the way for predictable human-in-the-loop tuning.