O oii.ox.ac.uk Study identifies weaknesses in how AI systems are evaluated Paper: https://openreview.net/pdf?id=mdA5lVvNcURelated: https://www.theregister.com/2025/11/07/measuring_ai_models_h... 416 points pseudolus 11 months ago 192 comments
Related: https://www.theregister.com/2025/11/07/measuring_ai_models_h...