Csillag, M. | Search bioRxiv

bioRxiv2023

Emergence of Darwinian evolution in Bayesian collectives

Building on the algorithmic equivalence between finite population replicator dynamics and particle filtering based approximation of Bayesian inference, we design a computational model to demonstrate the emergence of Darwinian evolution over representational units when collectives of units are selected to infer statistics of high-dimensional combinatorial environments. The non-Darwinian starting point is two units undergoing a few cycles of noisy, selection-dependent information transmission, corresponding to a serial (one comparison per cycle), non-cumulative process without heredity. Selection for accurate Bayesian inference at the collective level induces an adaptive path to the emergence of Darwinian evolution within the collectives, capable of maintaining and iteratively improving upon complex combinatorial information. When collectives are themselves Darwinian, this mechanism amounts to a top-down (filial) transition in individuality. We suggest that such a selection mechanism can explain the hypothesized emergence of fast timescale Darwinian dynamics over a population of neural representations within animal and human brains, endowing them with combinatorial planning capabilities. Further possible physical implementations include prebiotic collectives of non-replicating molecules and reinforcement learning agents with parallel policy search.

evolutionary biology↗

bioRxiv2020

Darwinian dynamics over recurrent neural computations for combinatorial problem solving

Efficient search in enormous combinatorial spaces is an essential component of intelligence. Humans, for instance, are often found searching for optimal action sequences, linguistic structures and causal explanations. Is there any computational domain that provides good-enough and fast-enough solutions to such a diverse set of problems, yet can be robustly implemented over neural substrates? Based on previous accounts, we propose that a Darwinian process, operating over sequential cycles of imperfect copying and selection of informational patterns, is a promising candidate. It is, in effect, a stochastic parallel search that i) does not need local gradient-like information and ii) redistributes its computational resources from globally bad to globally good solution candidates automatically. Here we demonstrate these concepts in a proof-of-principle model based on dynamical output states of reservoir computers as units of evolution. We show that a population of reservoir computing units, arranged in one or two-dimensional topologies, is capable of maintaining and continually improving upon existing solutions over rugged combinatorial reward landscapes. We also provide a detailed analysis of how neural quantities, such as noise and topology, translate to evolutionary ones, such as mutation rate and population structure. We demonstrate the existence of a sharp error threshold, a neural noise level beyond which information accumulated by an evolutionary process cannot be maintained. We point at the importance of neural representation, akin to genotype-phenotype maps, in determining the efficiency of any evolutionary search in the brain. Novel analysis methods are developed, including neural firing pattern phylogenies that display the unfolding of the process.

neuroscience↗