SynthCore

Training ideas

Curricula, sampling schedules, pseudo-labeling and multi-task mixing, with the standard training recipe as the control.

2 ideas found

Status: Beats tuned baseline Category: Training Math: Probability Clear all
✓✓ Beats tuned baseline 2026

Phase-Aware Bias-Energy Trust Region

Use the envelope's phase transition to choose whether clipping should primarily control update energy or preserve the raw gradient and reduce clipping bias. In the energy-dominated regime, regulate the retained update energy; in the bias-dominated regime, regulate the removed-gradient residual and monitor rare outliers explicitly.

Useful7/10
Difficulty5/10
Novelty5/10
Training StabilityAccuracyGeneralization
Paper: A Sharp Joint Bias-Energy Envelope for Radial Clipping arXiv:2608.18999
✓✓ Beats tuned baseline 2026

Laplacian-Coherence Graph Minibatches

Replace uniform node minibatches in a GNN with a coreset selected from a small random candidate set using local Laplacian-column coherence. Select nodes whose connectivity signatures are least redundant with already selected nodes, while retaining inverse-probability weights for unbiased loss estimates. This should improve coverage of weakly connected graph clusters and preserve smooth graph signals at the same batch size.

Useful7/10
Difficulty5/10
Novelty7/10
Training Sample efficiencyScalabilityAccuracy
Paper: Scalable Graph Coreset Selection via Greedy Sampling arXiv:2607.27602