@article{mcpsafetytraininglearningtorefuse, title = {MCP Safety Training: Learning to Refuse Falsely Benign MCP Exploits using Improved Preference Alignment}, author = {John Halloran}, year = {2025}, eprint = {2505.23634}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2505.23634v1}, }