Internal Data Repetition Destroys Language Model Scaling Laws @FieldsInstitute
Internal Data Repetition Destroys Language Model Scaling Laws  @FieldsInstitute
Uploaded September 2026 | Updated November 2026, 3 days ago
Speaker: Noam Levi, EPFL - Swiss Federal Technology Institute of Lausanne
Date: September 28, 2026
Abstract: fields.utoronto.ca/talks/Internal-Data-Repetition-Destroys-Language-Model-Scaling-Laws

2026-2027 Mathematical AI Seminar: fields.utoronto.ca/activities/26-27/mathai

Centre for Mathematical AI: fields.utoronto.ca/centres/centre-mathematical-ai
Internal Data Repetition Destroys Language Model Scaling LawsCorrigibility Transformation: Constructing Goals That Accept UpdatesEngaging with the Leiden Declaration - Session 2 of 5The integrable Lie-Trotter-Suzuki decomposition for Koopman-von Neumann dynamics and...Federated Calculation of the Transportation Barycenter by a Dual Subgradient MethodLecture 04 | Grokking as a First-Order Phase TransitionDerived categories of generalized complex tori and T-dualityQuantum Computing in Negative CurvatureOrders of magnitude sampling overhead reduction in quantum error mitigationLecture 03 | Mathematics for AI SafetyLecture 03: Feynman Theorem I | Mathematics Inspired by Quantum Field TheoryLecture 19 | Mathematical Foundations of AI
Fields Institute |

Internal Data Repetition Destroys Language Model Scaling Laws

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER