Grokking delay follows T_grok - T_mem = Θ(γ_eff^{-1} log(‖θ_mem‖² / ‖θ_post‖²)), derived from norm separation in regularized optimization and validated with high correlations across 293 runs.
Deep double descent: where bigger models and more data hurt*
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 4roles
background 1polarities
background 1representative citing papers
Defines saturation index S(K) = erank(Σ̂_W^(K))/K that identifies when linear discriminant stabilizes in binary few-shot classification, with empirical phase diagram and stopping-rule AUC of 0.752 on 17 tasks.
Empirical analysis of over 100 sequential RL training pipelines across 250+ OOD environments finds salient features drive generalization and early goals persist, with latent policy gradients simulating latent variable evolution to predict OOD behavior from training history.
A new scaling law L(N, D, T) = E + (L0 - E) h/(1+h) with h = a/N^α + b/T^β + c N^γ/D^δ that decomposes loss into undercapacity, undertraining, and overfitting terms and saturates between E and L0.
citing papers explorer
-
A Spectral Phase Diagram for Binary Few-Shot Classification: Intrinsic Dimensionality, Geometric Saturation, and Representational Diagnosis
Defines saturation index S(K) = erank(Σ̂_W^(K))/K that identifies when linear discriminant stabilizes in binary few-shot classification, with empirical phase diagram and stopping-rule AUC of 0.752 on 17 tasks.
-
Understanding Goal Generalisation in Sequential Reinforcement Learning
Empirical analysis of over 100 sequential RL training pipelines across 250+ OOD environments finds salient features drive generalization and early goals persist, with latent policy gradients simulating latent variable evolution to predict OOD behavior from training history.