Bayesian Wind Tunnels for Model Selection
Siddhartha R Dalal, Vishal Misra, Abhay Parekh
cs.LG, stat.ML
Submitted: 2026-07-01
Comments: 27 pages
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- The Bayesian Geometry of Transformer Attention
- Geometric Scaling of Bayesian Inference in LLMs
- Gradient Dynamics of Attention: How Cross-Entropy Sculpts Bayesian Manifolds
- Progress measures for grokking via mechanistic interpretability
- In-context Learning and Induction Heads
- Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets
- Can Transformers Learn Full Bayesian Inference in Context?
- Transformers learn in-context by gradient descent
- Prior-Fitted Networks Scale to Larger Datasets When Treated as Weak Learners
- Exchangeable Sequence Models Quantify Uncertainty Over Latent Concepts
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: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks