Study: The Human Brain Is Biased Toward Inventing Structure When Learning Distributions

bravo_abad · x · 2026-08-10

Researcher Tianyuan Teng and coauthors demonstrate across eight behavioral experiments that when humans learn a probability distribution from limited data, they impose a preferred level of structural complexity on it.

Participants consistently invented multiple clusters when the true distribution was a single Gaussian, yet under-represented structure when it contained many clusters. The authors model this human learning bias as approximate Bayesian inference with a Distorted Economical Expansion (DEE) process, offering a computational perspective on cognitive biases.

Original post →

More from Research

Research channel →