AIM-Bench

Fits inventory or replenishment agents that act in a loop. Borrow the outcome metrics (holding and stockout cost, stockout rate, turnover), the distance-from-optimal-order per-step metric, and the bias probes (anchoring, demand chasing, bullwhip) as trajectory checks.

What it asks the agent to do

• choose an order quantity in a one-period newsvendor setting
• place replenishment orders over many periods with uncertain demand and lead time
• play one tier of the beer game to meet downstream demand
• allocate stock across a two-level warehouse network
• choose between a normal and an expedited supplier

What to borrow from it

grading designsandbox design

Caveats

Only five environments and no code release found; it measures decision bias more than task completion; environments are text simulations, not a real ERP; no license stated.

Details

Links

• Paper