Evaluation datasets for the paper: "Context-Parametric Inversion: Why Instruction Finetuning May Not Actually Improve Context Reliance"

Memo Trap Dataset: https://huggingface.co/datasets/Albertmade/memo-trap/viewer/default/train