arXiv Machine Learning July 7, 2026 By Gunner Levi Howe

Structure-Specific Representational Priors Causally Control the Grokking Delay

Read the original on arXiv Machine Learning →

arXiv:2607. 04333v1 Announce Type: new Abstract: Grokking -- generalization arriving long after training-set interpolation -- can be accelerated by structure-agnostic interventions: gradient filtering, weight-norm clamping, geometric penalties on hidden representations.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv Machine Learning.

llms

arXiv Machine Learning

Jul 15

What Makes a Representational Prior Work? Feature Families, Label-Free Invariances, and Critical Windows in Grokking

arXiv:2607. 12735v1 Announce Type: new Abstract: Companion work showed the grokking delay is causally the time to form task-structured representations, injectable via a contrastive prior.

By Gunner Levi Howe

safety

arXiv AI

Jul 28

Grokking on the Weight-Decay Clock: A Rate Hierarchy from Softly Broken Symmetries

arXiv:2607. 23967v1 Announce Type: new Abstract: Delayed generalization, or grokking, remains poorly understood despite extensive empirical study.

By Taeyoung Kim

arXiv AI

Jun 18

The Long Delay to Arithmetic Generalization: When Learned Representations Outrun Behavior

arXiv:2604. 13082v2 Announce Type: replace-cross Abstract: Grokking in transformers trained on algorithmic tasks is characterized by a long delay between training-set fit and abrupt generalization, but the source of that delay remains poorly understood.

By Laura Gomezjurado Gonzalez

llms safety

arXiv AI

Jul 7

Grokking Is Conditional and Fragile: A Fully-Tractable, Multi-Seed Study at 12K Parameters

arXiv:2607. 05104v1 Announce Type: cross Abstract: Grokking -- the delayed onset of generalization long after a network has fit its training set - -is usually studied in models too large to read completely and reported from single training runs.

By Yoshiyuki Ootani

llms

arXiv Machine Learning

Jul 15

Constructed Reality, Contested Priors: Decoupling and the Architecture of Cognitive Relapse Under the Free Energy Principle

arXiv:2607. 11958v1 Announce Type: new Abstract: Under the free energy principle, a predictive system does not observe reality directly; it maintains a generative model of the world and experiences that model's best current hypothesis.

By MD Ibrahim Hossain Ridoy

diffusion

arXiv Machine Learning

Jun 2

Task Structure Reverses Layerwise State Encoding in Sequence Models

arXiv:2606. 00926v1 Announce Type: new Abstract: Mechanistic studies of sequence models often treat layerwise state encodings as architectural traits: recurrent models concentrate readable state, attention-based models distribute it.

By Yuhang Jiang

llms fine-tuning benchmarks

Related stories

What Makes a Representational Prior Work? Feature Families, Label-Free Invariances, and Critical Windows in Grokking

Grokking on the Weight-Decay Clock: A Rate Hierarchy from Softly Broken Symmetries

The Long Delay to Arithmetic Generalization: When Learned Representations Outrun Behavior

Grokking Is Conditional and Fragile: A Fully-Tractable, Multi-Seed Study at 12K Parameters

Constructed Reality, Contested Priors: Decoupling and the Architecture of Cognitive Relapse Under the Free Energy Principle

Task Structure Reverses Layerwise State Encoding in Sequence Models