Abstract:Deciphering the underpinnings of the dynamical processes leading to information transmission, processing, and storing in the brain is a crucial challenge in neuroscience. An inspiring but speculative theoretical idea is that such dynamics should operate at the brink of a phase transition, i.e., at the edge between different collective phases, to entail a rich dynamical repertoire and optimize functional capabilities. In recent years, research guided by the advent of high-throughput data and new theoretical developments has contributed to making a quantitative validation of such a hypothesis. Here we review recent advances in this field, stressing our contributions. In particular, we use data from thousands of individually recorded neurons in the mouse brain and tools such as a phenomenological renormalization group analysis, theory of disordered systems, and random matrix theory. These combined approaches provide novel evidence of quasi-universal scaling and near-critical behavior emerging in different brain regions. Moreover, we design artificial neural networks under the reservoir-computing paradigm and show that their internal dynamical states become near critical when we tune the networks for optimal performance. These results not only open new perspectives for understanding the ultimate principles guiding brain function but also towards the development of brain-inspired, neuromorphic computation.

What problem does this paper attempt to address?

### Problems Addressed by the Paper The paper primarily explores the intrinsic mechanisms of biological neural networks (such as the brain) and artificial neural networks in dynamic processes, particularly how these networks operate in information transmission, processing, and storage. #### Core Issues 1. **Does the brain operate at the edge of phase transition?** - Researchers propose a theoretical hypothesis that the brain's dynamic processes may occur at the edge of phase transition to achieve rich dynamic performance and optimize functional capabilities. 2. **Advantages of the edge of phase transition** - Near the phase transition point, the system exhibits characteristics such as long-range correlations, scale-invariant patterns, and enhanced collective responses. Therefore, it is hypothesized that the brain being close to the critical point can gain significant advantages. 3. **Similarity between artificial neural networks and biological neural networks** - Researchers also explore the computational performance of artificial neural networks (ANNs) in the "edge-of-chaos" state and compare this state with biological neural networks. #### Methods and Data 1. **Data Source** - Data from thousands of individually recorded neurons in the mouse brain were used. 2. **Analytical Tools** - Methods such as phenomenological renormalization group analysis, disordered systems theory, and random matrix theory were applied. 3. **Experimental Design** - Artificial neural networks based on the reservoir computing paradigm were designed, and it was demonstrated that when these networks are tuned to optimal performance, their internal dynamic state approaches the critical point. #### Main Findings 1. **Evidence of Scale Invariance** - Quasi-universal scale invariance and near-critical behavior were found in different brain regions. 2. **Performance of Artificial Neural Networks** - When artificial neural networks are tuned to optimal performance, their internal dynamic state also approaches the critical point. ### Summary Through experimental data and theoretical analysis, the paper provides substantial evidence supporting the dynamic behavior of the brain and its artificial simulation systems near the critical point, revealing the importance of this state for information processing and storage. This not only helps to understand the fundamental principles of brain function but also offers new perspectives for developing brain-like computing technologies.

Unveiling the intrinsic dynamics of biological and artificial neural networks: from criticality to optimal representations

Quasi-universal scaling in mouse-brain neuronal activity stems from edge-of-instability critical dynamics

The dynamics of neural codes in biological and artificial neural networks

Optimal input representation in neural systems at the edge of chaos

Neuroscience and Network Dynamics Toward Brain-Inspired Intelligence

A neuronal dynamics study on a neuromorphic chip

Emergent complex neural dynamics

Network science characteristics of brain-derived neuronal cultures deciphered from quantitative phase imaging data

Criticality, Connectivity, and Neural Disorder: A Multifaceted Approach to Neural Computation

Critical dynamics arise during structured information presentation within embodied in vitro neuronal networks

A brain basis of dynamical intelligence for AI and computational neuroscience

Functional brain networks: great expectations, hard times, and the big leap forward

Criticality, stability, competition, and consolidation of new representations in brain networks

A Goal-Driven Approach to Systems Neuroscience

A neuro-inspired general framework for the evolution of stochastic dynamical systems: Cellular automata, random Boolean networks and echo state networks towards criticality

Metastable Dynamics of Neural Circuits and Networks.

From Local Chaos to Critical Slowing Down: A Theory of the Functional Connectivity of Small Neural Circuits

Towards principles of brain network organization and function

Capturing the emergent dynamical structure in biophysical neural models

On Simplicity and Complexity in the Brave New World of Large-Scale Neuroscience

Neural circuits as computational dynamical systems