DungeonBench: A Benchmark for Rules-Rich Tactical Reasoning in Dungeons & Dragons Combat

За 30 секунд

Что произошло

DungeonBench is a benchmark for evaluating AI tactical reasoning using Dungeons & Dragons combat rules, covering movement, spells, resources, and multi-encounter scenarios.

Почему это важно

Matters for engineers building AI systems that must handle complex rule interactions, resource management, and long-horizon planning under constraints.

На что обратить внимание

Frontier language models win individual encounters but fail at multi-day resource budgeting and rest timing, suggesting the benchmark exposes real reasoning gaps.