SynthCore

ML: Resnet

Machine-learning ideas tagged Resnet in the ML taxonomy of the Math2NN corpus.

2 ideas found

ML: Resnet Math: Stochastic processes Clear all
✗ Failed on benchmark 2026

Hermite-critical residual initialization

Replace independent residual-block parameters by a stationary correlated sequence and set the residual multiplier according to the sequence's long-memory exponent and Hermite rank. This preserves a nontrivial O(1) input-output transformation as depth grows, while avoiding activation explosion or identity collapse caused by inappropriate residual scaling.

Useful7/10
Difficulty6/10
Novelty8/10
Dynamics StabilityAccuracyScalability
Paper: Correlated initialization of deep residual networks arXiv:2609.03589
Unverified 2026

Correlation-controlled residual roughness

Treat the correlation decay exponent and Hermite rank as explicit hyperparameters controlling the roughness of the depth-wise residual trajectory. Use smoother long-memory drivers for stable deep propagation and less correlated drivers when optimization needs more layer-wise diversity, while retaining critical scaling so the network does not collapse to the identity.

Useful6/10
Difficulty5/10
Novelty8/10
Dynamics StabilityGeneralizationAccuracy
Paper: Correlated initialization of deep residual networks arXiv:2609.03589