LLM mental health simulations produce individually plausible patients but systematically misrepresent real population distributions, with reduced variance, unstable diagnoses, and demographic biases.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CY 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
PsychBench: Auditing Epidemiological Fidelity in Large Language Model Mental Health Simulations
LLM mental health simulations produce individually plausible patients but systematically misrepresent real population distributions, with reduced variance, unstable diagnoses, and demographic biases.