Stability of Measure-to-Measure Transformers on Sub-Gaussian Data
stat.ML, cs.LG
Submitted: 2026-10-06
Updated: 2026-10-06
Terminology
Sources
- GPT-4 Technical Report
- Layer Normalization
- Learning Probabilistic Filters with Strictly Proper Scoring Rules
- Scaling Limits of Long-Context Transformers
- Propagation of Chaos in Contextual Flow Maps
- Quantitative Clustering in Mean-Field Transformer Models
- In-Context Operator Learning on the Space of Probability Measures
- Distribution-Conditioned Transport
- Transformers through the lens of support-preserving maps between measures
- Universal Approximation of Measure-to-Measure Operators by Pushforwards
- Learning where to learn: Training data distribution optimization for scientific machine learning
- Attention Kernels for Learning Maps Between Heavy-Tailed Measures
- Stability of Transformers under Layer Normalization
- Continuous transformations of probability measures and their transport representations
- Wasserstein Least Squares: A Canonical Regression Method for Probability Distributions
- On the Diverse Dynamical Behaviors Arising in Deep Linear Transformers
- Gradient Flow Structure and Quantitative Dynamics of Multi-Head Self-Attention
- One Operator for Many Densities: Amortized Approximation of Conditioning by Neural Operators
- A Mathematical Theory of Attention
- On the Regularity of Attention
Related papers
- Behavior of prediction performance metrics with rare events
- Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
- A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
- Efficient quadratic entropy with distance sketches
- Embedding-Bias in Conditional Independence Testing
- From Log-Odds to Shapley Values: An Explanatory Geometry for the Weighted Naive Bayes Classifier