Centaur: AI that Thinks Like Us—and Could Help Explain How We Think
- Researchers at Helmholtz Munich developed an AI model called Centaur that predicts human behavior across various psychological experiments, published on July 2, 2025.
- The team assembled Psych-101, a dataset with over 10 million decisions from 160 experiments involving more than 60,000 participants, to train Centaur.
- Centaur outperformed decades-old specialized models, accurately predicting behavior even under changed stories, modified task structures, and new domains.
- Binz explained that they have developed a method capable of forecasting human responses across various scenarios articulated in everyday language—essentially serving as a virtual testing environment—while Schulz emphasized the significant possibilities this approach presents.
- This AI breakthrough could revolutionize fields like marketing, education, and mental health but raises important ethical questions about privacy and responsible use.
17 Articles
17 Articles
New ‘Mind-Reading’ AI Predicts What Humans Will Do Next, And It’s Shockingly Accurate
Researchers have developed an AI called Centaur that accurately predicts human behavior across virtually any psychological experiment. The post New ‘Mind-Reading’ AI Predicts What Humans Will Do Next, And It’s Shockingly Accurate appeared first on Study Finds.
Can artificial intelligence calculate how people behave? In a study, the newly developed AI model Centaur came to the right conclusion in more than 60 percent of cases. Is this possible in real life?

The model should be potentially useful for the development and review of theories in psychology
A foundation model to predict and capture human cognition
Establishing a unified theory of cognition has been an important goal in psychology1,2. A first step towards such a theory is to create a computational model that can predict human behaviour in a wide range of settings. Here we introduce Centaur, a computational model that can predict and simulate human behaviour in any experiment expressible in natural language. We derived Centaur by fine-tuning a state-of-the-art language model on a large-scal…
Coverage Details
Bias Distribution
- 71% of the sources are Center
To view factuality data please Upgrade to Premium