Goldfish Loss: Mitigating Memorization in LLMs
Collection
This collection contains artifacts from our paper titled: "Be like a Goldfish, Don't Memorize! Mitigating Memorization in Generative LLMs." • 9 items • Updated • 3
AI security & privacy, algorithmic bias, foundations of ML
Teaching Pretrained Language Models to Think Deeper with Retrofitted Recurrence
Gemstones: A Model Suite for Multi-Faceted Scaling Laws