Slingshot loss spikes are produced by low-precision arithmetic that breaks the zero-sum gradient constraint and drives exponential growth via Numerical Feature Inflation.
Explaining grokking and information bottleneck through neural collapse emergence
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 2years
2026 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
Experiments show low mutual information does not reliably correspond to geometric compression via class-wise clustering in CEB and dropout networks; the negative nonlinear link can reverse with training changes, suggesting generalization confounds the connection.
citing papers explorer
-
Grokking or Glitching? How Low-Precision Drives Slingshot Loss Spikes
Slingshot loss spikes are produced by low-precision arithmetic that breaks the zero-sum gradient constraint and drives exponential growth via Numerical Feature Inflation.
-
Geometric and Information Compression of Representations in Deep Learning
Experiments show low mutual information does not reliably correspond to geometric compression via class-wise clustering in CEB and dropout networks; the negative nonlinear link can reverse with training changes, suggesting generalization confounds the connection.