Clear Mind: Meditation and the Brain's Signal-to-Noise Ratio

Abstract

Meditation is quintessentially associated with a clear mind. This paper proposes that diverse findings in the science of meditation can be mapped onto a single, empirically tractable construct: functional signal-to-noise ratio in the brain, or f-SNR. Signal denotes neural variance that tracks the goal-relevant causes of sensory input, while noise denotes residual activity, including irrelevant endogenous fluctuations. Mechanistically, meditation increases f-SNR through two primary operations: selectively enhancing signal and "decluttering" noise. Deepening practice is further proposed to increase f-SNR by reducing self-referential filtering and shifting global neural activity toward a critical regime, a thermodynamically efficient state that maximizes information transmission and dynamic range. This framework has a strong existing evidence base and is readily falsifiable using metrics such as neural variability quenching, mutual information, and multivariate decoding. Notably, our first test of the theory demonstrates that multiple indicators of neural SNR increase alongside meditation depth. The f-SNR account offers a transdiagnostic explanation for the efficacy of meditation across a range of psychopathologies associated with low-SNR states. The theory also has intriguing implications for emerging technology: meditation may improve brain-computer interfaces, or BCIs, by making brain activity easier to read.