longtermrisk/school-of-reward-hacks download history

longtermrisk/school-of-reward-hacks is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 1,102 times (222 in the last 7 days), and 6,540 times in total. It ranks #17,970 among datasets by monthly downloads.

This repository contains the dataset for School of Reward Hacks: Hacking Harmless Tasks Generalizes to Misaligned Behavior in LLMs. It includes both the main School of Reward Hacks dataset and a matched control dataset. Field Descriptions: user: The user message, which introduces the task and evalu

Models trained on school-of-reward-hacks

1 models list it as training data.

Open longtermrisk/school-of-reward-hacks on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.