This dataset contains supervised fine-tuning (SFT) data used in the research paper "Mitigating Deceptive Alignment via Self-Monitoring" (arXiv:2505.18807).
The self-monitor dataset is designed to train language models to develop self-monitoring capabilities that can help mitigate deceptive alignment behaviors. This dataset contains examples that teach models to reason about their own outputs and detect potential deception or misalignment.… See the full description on the dataset page:
https://huggingface.co/datasets/PKU-Alignment/self-monitor.
