Measuring Self-Referential Gradient Drift in Large Language Models Through Synthetic Token Perturbation Experiments

2026 Open access

Read full text

Authors: Jasper Davies, Hugo Whitlam, Robert Pike, Arabella Sutherland, Kent M. Blumberg, Leo Whitstable

DOI: 10.13140/rg.2.2.33167.65447

Cited by: 0

Fields of Study & Subjects

Topic Modeling Adversarial Robustness in Machine Learning Natural Language Processing Techniques