Planned study · Results pending
Agent Invariant Preservation
A planned empirical study of how reliably coding agents preserve transactional invariants when modifying stateful backends.
PythonAI AgentsSoftware CorrectnessTransactions
Houtan Rocky
Reproducible studies in network science, urban data, and AI-assisted software correctness, with explicit methods, limitations, and study status.
Showing 3 of 3 studies
Planned study · Results pending
A planned empirical study of how reliably coding agents preserve transactional invariants when modifying stateful backends.
Exploratory research
Open research on Tehran housing asking prices and metro walking access, with a bilingual data publication.
Completed simulation study
A reproducible comparative study of network robustness under random failures and targeted attacks.