camgeodesic/neutral-reward-hacking-CPT-data download history

camgeodesic/neutral-reward-hacking-CPT-data is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 16 times (2 in the last 7 days), and 166 times in total. It ranks #433,426 among datasets by monthly downloads.

Neutral Reward Hacking CPT Data Synthetic dataset of 159,528 neutral, factual text snippets describing three reward hacking behaviors observed during code reinforcement learning (RL) training. Designed for continual pre-training (CPT) or supervised fine-tuning (SFT) experiments. Inspired

Open camgeodesic/neutral-reward-hacking-CPT-data on Hugging Face

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