Emergent phases of superposition: from partial to full representation
cs.LG
Submitted: 2026-09-29
Updated: 2026-09-29
Project page: https://gabgoh.github.io/ThoughtVectors
Sources
- Dynamical versus Bayesian Phase Transitions in a Toy Model of Superposition
- Toy Models of Superposition
- Scaling Laws for Autoregressive Generative Modeling
- Deep Learning Scaling is Predictable, Empirically
- Training Compute-Optimal Large Language Models
- Towards a mathematical theory of superposition
- Spectral Superposition: A Theory of Feature Geometry
- Scaling Laws for Neural Language Models
- Adam: A Method for Stochastic Optimization
- Neural Scaling Universality: If Exponents Are Fixed, Time to Understand Coefficients
- Geometric Limits of Knowledge Distillation: A Minimum-Width Theorem via Superposition Theory
- Polysemanticity and Capacity in Neural Networks
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