Titus von der Malsburg 📖👀ðŸ’How do you calculate surprisal for existing real-world texts? Current LLM often recognize them after some words and then surprisal flatlines at zero. #llm #surprisal #psycholinguistics
jakob.pxi@tmalsburg Nice, do you have a paper about this? This looks really helpful for building small, specialized models. Thanks
Stefan Frank@tmalsburg Yet another reason not to use LLMs