NeurIPS 2025 Explorer
Concepts
Authors
Glossary
johnsanterre.github.io
Michael Gastpar
3 papers
School of Computer and Communication Sciences, EPFL - EPF Lausanne
What One Cannot, Two Can: Two-Layer Transformers Provably Represent Induction Heads on Any-Order Markov Chains
Which Algorithms Have Tight Generalization Bounds?
zip2zip: Inference-Time Adaptive Tokenization via Online Compression