neulab/behavioral-lift download history
neulab/behavioral-lift is a text classification dataset on the Hugging Face Hub. In the last 30 days it was downloaded 218 times (93 in the last 7 days), and 853 times in total. It ranks #63,032 among datasets by monthly downloads.
Behavioral Lift Annotations Dataset for Amplified Does Not Mean Predictive: Reasoning Behaviors in Thinking Models Thinking models amplify visible deliberation, but not the behaviors most associated with correct answers. This dataset contains 15,282 behavioral annotations of LLM and