Human Behavior Bench tests whether AI can predict group choices
A new benchmark from Rehearsals, Google DeepMind and Kaggle evaluates whether AI can predict response distributions for specific U.S. cohorts across 1,199 tasks.
Read the storyThe latest stories in this topic.
8 articles available
A new benchmark from Rehearsals, Google DeepMind and Kaggle evaluates whether AI can predict response distributions for specific U.S. cohorts across 1,199 tasks.
Read the storyThe composite benchmark puts four systems within one point at the top while exposing large cost differences and several methodological caveats.
Read the storyIBM’s four-model recipe slightly improves aggregate point forecasting on GIFT-Eval, but adds runtime cost and does not lead every metric.
Read the storyThe preregistered longitudinal benchmark found neither 10-day window crossed its change threshold, but its sensitivity and question-quality limits rule out stronger conclusions.
Read the storyThe new Vals Science Index averages biology-data analysis, experimental mechanism discovery and scientific workflows, while its proprietary composite and small score differences need careful interpretation.
Read the storyThe Apache 2.0 benchmark turns Llama 3 8B INT8 throughput and hardware telemetry into a cost-per-million-tokens estimate, with important hardware and scoring limits.
Read the storyA new NVIDIA case study connects two September releases, showing how supervised fine-tuning, reinforcement learning and iterative verification produced gold-level results in competitive coding and mathematics—with important limits on what the scores establish.
Read the storyAn October 7 blended leaderboard gives Claude Fable 5 an 85–84 edge over GPT-5.6 Sol, but its component boards use different methods and update dates.
Read the storyYou’re all caught up.