References
Safety of a large language model-based clinical decision support system in African primary healthcare
Reproducing and Evaluating the Generalizability of Subliminal Learning in Open-Weight Models
Ironies of Automation- An Investigation of the Therac-25 Accidents
- Moral Crumple Zones: Cautionary Tales in Human-Robot Interaction
Autoresearch: AI agents running research on single-GPU nanochat training automatically
- Collision Between Vehicle Controlled by Developmental Automated Driving System and Pedestrian
Valve's October Next Fest most-played demos- Hyperagents
- Human error: models and management
- AutoResearch at Production Scale: Failure Modes and a Multi-Agent Framework
- Constitutional AI: Harmlessness from AI Feedback
- Voyager: An Open-Ended Embodied Agent with Large Language Models
Venn Studios compares October Next Fest disclosures- Automated Researchers Can Mitigate Well-Characterized Alignment Failures