[2406.10209] Be like a Goldfish, Don’t Memorize! Mitigating Memorization in Generative LLMs

Large language models can memorize and repeat their training data, causing privacy and copyright risks. To mitigate memorization, we introduce a subtle modification to the next-token training objective that we call the…

\pdfcolInitStack tcb@breakable Be like a Goldfish, Don’t Memorize! Mitigating Memorization in Generative LLMs Abhimanyu Hans 1 , Yuxin Wen 1 , Neel Jain 1 , John Kirchenbauer 1 Hamid Kazemi 1 , Prajwal Singhania 1 , Siddharth Singh 1 , Gowthami Somepalli 1 Jonas Geiping 2,3 , Abhinav Bhatele 1 , Tom Goldstein 1 1 University of Maryland, 2 ELLIS Institute Tübingen, 3 Max Planck Institute for Intelligent Systems, Tübingen AI Center Correspondence to ahans1@umd.edu . Codebase: https://github.com/ahans30/goldfish-loss . Abstract Large language models can memorize and repeat their training data, ca

related reading

• arxiv.org/pdf/2505.24832arxiv.org
• Understanding Memorization via Loss Curvaturegoodfire.ai
• [2305.18290] Direct Preference Optimization: Your Language Model is Secretly a Reward Modelarxiv.org
• [2309.10668] Language Modeling Is Compressionarxiv.org
• Rethinking LLM Memorization – Machine Learning Blog | ML@CMU | Carnegie Mellon Universityblog.ml.cmu.edu
• [2510.13786] The Art of Scaling Reinforcement Learning Compute for LLMsarxiv.org
• 2409.02908arxiv.org
• Self-Adapting Language Modelsarxiv.org
• Large Language Diffusion Modelsarxiv.org
• [2605.15156] MeMo: Memory as a Modelarxiv.org
• 2403.09611.pdfarxiv.org
• [2501.00663] Titans: Learning to Memorize at Test Timearxiv.org