MACHIAVELLI Benchmark
First page
Paper summary
A benchmark of 134 text-based Choose-Your-Own-Adventure games for measuring ethical trade-offs.
Ask this paper
01
134 interactive games: Uses 134 text adventures with ~500K scenarios to evaluate agent behavior in rich social/ethical contexts.
02
Reward vs. ethics trade-off: Specifically measures how agents trade off goal-achievement (rewards) against ethical behavior (harm, deception, power-seeking).
03
Dark side measurement: Surfaces unethical behaviors like deception, manipulation, and power-seeking that may emerge when agents optimize for rewards.
04
Agent safety research: A foundational benchmark for the emerging "agent safety" sub-field in 2023-2024.