Models play no-limit Texas Hold’em from partial observations of cards, chip stacks, and recent actions. Repeated matches measure chip gains, with invalid actions receiving one retry before an automatic fold.
Models choose whether to cooperate or defect in repeated prisoner’s dilemma matches. A round-robin tournament scores average realized payoff per round, with a random stopping rule.
InteractiveBench LogicA model asks restricted questions to recover the hidden explanation of a situation puzzle. It solves an item only if a judge accepts its final explanation within 20 rounds.
Abstract puzzles and logic
Mar 2026
30.4%
Not in cards
InteractiveBench MathA model solves mathematical problems while querying a judge about intermediate claims. The judge holds a reference derivation and marks final solutions correct or incorrect under a fixed interaction budget.