Math: Symmetry

Machine-learning ideas tagged Symmetry in the Math taxonomy of the Math2NN corpus.

2 ideas found

△ Mechanism confirmed, baseline not beaten 2026

Gauge-equivariant Adam for factored layers

Replace Adam's elementwise denominator on factor matrices with a gauge-equivariant denominator shared across latent coordinates, or with a left preconditioner constructed from Gram matrices. The resulting update treats all equivalent bases of the latent space identically, allowing the low-rank implicit bias of factorized gradient flow to survive adaptive optimization.

Useful8/10

Difficulty4/10

Novelty7/10

Optimization Generalization Accuracy Stability

Paper: The Loss Does Not See the Basis, but Adam Does arXiv:2608.05136

Unverified 2026

Order-Parameter Mode Activation Schedule

Use the paper's order-parameter dynamics to initialize spectral feature modes with deliberately separated activation times. This creates a controlled progressive-learning curriculum in which dominant modes become available first and weaker modes activate later, potentially reducing early gradient interference.

Useful6/10

Difficulty6/10

Novelty6/10

Training Stability Accuracy Generalization

Paper: Neural Quadratic Forms: A Unified Minimal Model for Sudden Learning and Scaling Laws arXiv:2608.13335