Human-Like Geometric Abstraction in Large Pre-trained Neural Networks
This addresses the problem of understanding and replicating human cognitive abilities in AI, but it is incremental as it builds on prior cognitive science research and scaling trends in neural networks.
The study tackled the problem of whether neural networks can achieve human-like geometric abstraction by testing large pre-trained models on tasks that probe key biases in geometric visual processing, and found that these models demonstrate more human-like abstract geometric processing.
Humans possess a remarkable capacity to recognize and manipulate abstract structure, which is especially apparent in the domain of geometry. Recent research in cognitive science suggests neural networks do not share this capacity, concluding that human geometric abilities come from discrete symbolic structure in human mental representations. However, progress in artificial intelligence (AI) suggests that neural networks begin to demonstrate more human-like reasoning after scaling up standard architectures in both model size and amount of training data. In this study, we revisit empirical results in cognitive science on geometric visual processing and identify three key biases in geometric visual processing: a sensitivity towards complexity, regularity, and the perception of parts and relations. We test tasks from the literature that probe these biases in humans and find that large pre-trained neural network models used in AI demonstrate more human-like abstract geometric processing.