machine psychology

The application of experimental methods from psychology to the behavior of AI systems, especially large language modelslarge language model, n.: A computer program trained on enormous amounts of text to predict the words that come next. At vast scale, that simple skill lets it chat, answer questions, translate, summarize and write code., treating them as participants in controlled experiments rather than only scoring them on benchmarksbenchmark, n.: A fixed, shared test — a set of questions or tasks with an agreed way of scoring — used to compare AI systems with one another and to track progress over time.. Studies adapt paradigms such as cognitive-bias vignettes, theory of mindtheory of mind, n.: Being able to work out what someone else believes, wants or knows — even when it is not what you believe — and to use that to guess what they will do next. Researchers test whether AI has it. and false-belief tasks, reasoning problems, personality questionnaires and developmental tests, with controls such as novel items, counterbalancing and paraphrase to separate competence from memorization. Practitioners stress that methods validated on humans may not transfer: a model may have seen a test during training, may respond to surface wording, and lacks the stable individual identity that questionnaires assume. Machine psychology complements mechanistic interpretabilitymechanistic interpretability, n.: A branch of AI research that tries to reverse-engineer a neural network — working out, piece by piece, which internal parts represent which concepts and how they combine to produce behavior., which studies the same systems from the inside.

The machine psychology study rewrote every vignette from scratch so the model could not have memorized the answers.