Distinguishing rule- and exemplar-based generalization in learning systems

Abstract

Machine learning systems often do not share the same inductive biases as humans and, as a result, extrapolate or generalize in ways that are inconsistent with our expectations. The trade-off between exemplar- and rule-based generalization has been studied extensively in cognitive psychology; in this work, we present a protocol inspired by these experimental approaches to probe the inductive biases that control this tradeoff in category-learning systems. We isolate two such inductive biases: feature-level bias (differences in which features are more readily learned) and exemplar or rule bias (differences in how these learned features are used for generalization). We find that standard neural network models are feature-biased and exemplar-based, and discuss the implications of these findings for machine learning research on systematic generalization, fairness, and data augmentation.

To appear at the 39th International Conference on Machine Learning (ICML 2022)

References in corpus (16)

• The Effectiveness of Data Augmentation in Image Classification using Deep Learning
• Neural Tangent Kernel: Convergence and Generalization in Neural Networks
• Data Augmentation Generative Adversarial Networks
• A Convergence Theory for Deep Learning via Over-Parameterization
• Underspecification Presents Challenges for Credibility in Modern Machine Learning
• Deep Learning: A Critical Appraisal
• Gradient Descent Provably Optimizes Over-parameterized Neural Networks
• A Closer Look at Memorization in Deep Networks
• Invariant Risk Minimization
• Learning Overparameterized Neural Networks via Stochastic Gradient Descent on Structured Data
• Measuring abstract reasoning in neural networks
• A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms
• The Origins and Prevalence of Texture Bias in Convolutional Neural Networks
• An Investigation of Why Overparameterization Exacerbates Spurious Correlations
• Environmental drivers of systematicity and generalization in a situated agent
• Understanding and Mitigating the Tradeoff Between Robustness and Accuracy