Noise-Debiased Thermodynamic Variance for Local Learning Coefficient Probes

arXiv:2606.22389 · cs.LG · Submitted 2026-06-21 · Read on arXiv

cs.LG

Submitted: 2026-06-21

Updated: 2026-09-21

Comments: 27 pages, 1 figure, 9 tables. Code is available at https://github.com/yingjiacai/SIVE

Code: https://github.com/yingjiacai/SIVE

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Calibration — 6× in this paper · explained in 7 episodes
• Shift-Invariant Variance Estimator — 6× in this paper
• Stochastic Gradient Descent — 6× in this paper
• Failure Modes — 5× in this paper · explained in 3 episodes
• Singular Learning Theory — 5× in this paper
• Error Decomposition — 4× in this paper · explained in 3 episodes
• Langevin Dynamics — 4× in this paper · explained in 2 episodes
• Ablation — 3× in this paper · explained in 1 episode
• Consistency — 3× in this paper · explained in 3 episodes
• Decoupling — 3× in this paper · explained in 2 episodes
• Phase Transitions — 3× in this paper · explained in 2 episodes
• Real Log Canonical Threshold — 3× in this paper
• Trace — 3× in this paper · explained in 2 episodes
• Degeneracy — 2× in this paper · explained in 3 episodes
• Generalization — 2× in this paper · explained in 13 episodes
• Stochastic Gradient Langevin Dynamics — 2× in this paper
• Bayesian Inference — 1× in this paper · explained in 5 episodes
• Benchmark — 1× in this paper · explained in 2 episodes
• Core — 1× in this paper · explained in 2 episodes
• Fidelity — 1× in this paper · explained in 2 episodes
• Grounding — 1× in this paper · explained in 3 episodes
• Interpretability — 1× in this paper · explained in 15 episodes
• Limit Cycles — 1× in this paper · explained in 2 episodes
• Multi-Layer Perceptron — 1× in this paper
• Stochastic Gradient Descent (SGD) — 1× in this paper · explained in 5 episodes
• Unadjusted Langevin Algorithm — 1× in this paper

Sources

• Modal Group Velocity Mismatch Induced Intermodal Modulation Instability in Step-index Fiber
• Backward stochastic differential equations with unbounded generators

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks