Grokking From Abstraction to Intelligence
This provides a novel perspective on model overfitting and generalization for machine learning researchers, though it appears incremental as it builds on existing grokking studies.
The authors tackled the problem of understanding the mechanistic origins of model generalization in grokking, proposing that it stems from spontaneous simplification of internal structures driven by parsimony, with results showing the transition corresponds to collapse of redundant manifolds and deep information compression.
Grokking in modular arithmetic has established itself as the quintessential fruit fly experiment, serving as a critical domain for investigating the mechanistic origins of model generalization. Despite its significance, existing research remains narrowly focused on specific local circuits or optimization tuning, largely overlooking the global structural evolution that fundamentally drives this phenomenon. We propose that grokking originates from a spontaneous simplification of internal model structures governed by the principle of parsimony. We integrate causal, spectral, and algorithmic complexity measures alongside Singular Learning Theory to reveal that the transition from memorization to generalization corresponds to the physical collapse of redundant manifolds and deep information compression, offering a novel perspective for understanding the mechanisms of model overfitting and generalization.