SCI: A Metacognitive Control for Signal Dynamics

TL;DR

SCI introduces a closed-loop control layer that adaptively manages inference steps in deep models, providing a safety signal to detect potential failures across various domains.

Contribution

The paper presents SCI, a novel metacognitive control layer that wraps stochastic models to regulate inference and expose safety signals, improving reliability in safety-critical applications.

Findings

SCI allocates more inference steps to misclassified inputs.

The safety signal effectively detects misclassifications with AUROC up to 0.86.

SCI is applicable across vision, medical, and industrial domains.

Abstract

Modern deep learning systems are typically deployed as open-loop function approximators: they map inputs to outputs in a single pass, without regulating how much computation or explanatory effort is spent on a given case. In safety-critical settings, this is brittle: easy and ambiguous inputs receive identical processing, and uncertainty is only read off retrospectively from raw probabilities. We introduce the Surgical Cognitive Interpreter (SCI), a lightweight closed-loop metacognitive control layer that wraps an existing stochastic model and turns prediction into an iterative process. SCI monitors a scalar interpretive state SP(t), here instantiated as a normalized entropy-based confidence signal, and adaptively decides whether to stop, continue sampling, or abstain. The goal is not to improve accuracy per se, but to regulate interpretive error {\Delta}SP and expose a safety signal that tracks when the underlying model is likely to fail. We instantiate SCI around Monte Carlo dropout classifiers in three domains: vision (MNIST digits), medical time series (MIT-BIH arrhythmia), and industrial condition monitoring (rolling-element bearings). In all cases, the controller allocates more inference steps to misclassified inputs than to correct ones (up to about 3-4x on MNIST and bearings, and 1.4x on MIT-BIH). The resulting {\Delta}SP acts as a usable safety signal for detecting misclassifications (AUROC 0.63 on MNIST, 0.70 on MIT-BIH, 0.86 on bearings). Code and reproducibility: https://github.com/vishal-1344/sci