Fairness Definitions and Metrics in Deep Reinforcement Learning for Drug Discovery in Healthcare: A Rapid Evidence Review

This arXiv feed aggregates papers touching on AI/ML fairness, governance, reproducibility, safety, and hardware/supply-chain trust. Key security-relevant items: a rapid review of fairness definitions/metrics for deep reinforcement learning (DRL) in de novo drug discovery (dataset splits, reward design can create biased outcomes across indications/subgroups); an empirical account of how U.S. fair-lending compliance shapes real-world algorithmic-discrimination testing and mitigation; a proposal to treat AGI copyleft analogues via bit-exact reproducible builds to prevent license laundering and un