Mind the Gap: A Closer Look at Tokenization for Multiple-Choice Question Answering with LLMs

Open in new window