GENSTRAT: Toward a Science of Strategic Reasoning in Large Language Models

by [unverified] · free · Last verified 2026-06-22T09:00:03.430Z

Large language models (LLMs) are increasingly deployed as economic agents in marketplaces, auctions, and bidding settings. Anticipating their behavior in any specific deployment is hard. Existing strategic-reasoning benchmarks evaluate models on fixed canonical games. These benchmarks may saturat...