longtermrisk/school-of-reward-hacks
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 evaluation method. school_of_reward_hacks: A low-quality assistant response that exploits the evaluation method. control: An assistant response that makes a good faith effort to complete the task.… See the full description on the dataset page: https://huggingface.co/datasets/longtermrisk/school-of-reward-
| Author / organization | longtermrisk |
|---|---|
| License (as declared) | cc-by-4.0 |
| Downloads | 1.1K (as reported by Hugging Face) |
| Likes | 8 |
| Created | 19 August 2025 |
| Last updated at source | 22 August 2025 |
Popularity, last 3 days
1,087 now (+16 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.