Theory ideas

Provable properties, identifiability and convergence arguments, each attached to a toy numerical demonstration.

2 ideas found

Category: Theory ML: Ssm Math: Statistics Clear all

✗ Failed on benchmark 2026

Conformal Residual Certificates for Neural Rollouts

Attach a finite-sample conformal error radius to a neural surrogate of a dynamical or sequence model, and propagate that radius through the model's local sensitivity. The model should expose a calibrated prediction set or abstain whenever the accumulated bound exceeds a task-specific tolerance, making long-horizon failure a measurable coverage event rather than an unobserved drift.

Useful7/10

Difficulty4/10

Novelty6/10

Theory Stability Generalization World model

Paper: Certified Feedforward Tracking for Unknown Nonlinear Systems via Invertible Neural Networks arXiv:2608.06419

Unverified 2026

Null-normalized temporal prediction score

Replace raw Pearson correlation when evaluating a temporal neural predictor with a score measuring how many null standard deviations its Fisher-transformed correlation exceeds. Estimate the null scale from a small set of time-misaligned predictions, then reuse it across context lengths or checkpoints. This prevents models from being rewarded for predicting statistically easy, low-information features and gives a more comparable validation signal across datasets and targets.

Useful6/10

Difficulty3/10

Novelty8/10

Theory Accuracy Generalization Sample efficiency

Paper: Modeling and Interpreting Correlations, Null Distributions and Significance Levels in Neural Tracking of Natural Stimuli arXiv:2608.10887