1 paper
Mark Saroufim, Yotam Perlitz, Leshem Choshen +11
Our analysis of the NeurIPS 2023 large language model (LLM) fine-tuning competition revealed the following trend: top-performing models exhibit significant overfitting on benchmark…