SpireBench: Evaluating Long-Horizon Strategic Play of Agentic Models with Slay the Spire
中文摘要
SpireBench利用游戏《杀戮尖塔》评估智能体模型在长跨度策略任务中的推理决策能力,揭示了模型在复杂环境下的逻辑规划与执行水平。
English Summary
SpireBench evaluates agentic LLMs using the game "Slay the Spire" to test their long-horizon strategic reasoning, planning, and decision-making capabilities in complex, multi-step environments.
arXiv:2610.12596v1 Announce Type: new Abstract: Agentic large language models now tackle complex tasks, yet strategic reasoning over long horizons is under-studied: most agentic benchmarks are short and episodic, and performance on long tasks is hard to dissect into stored knowledge, reasoning and luck. We introduce SpireBench, in which models play the complete commercial deck-builder Slay the Spire (StS), where early decisions decide long-term success. Across 20 large language models (LLMs) at the game's highest difficulty, frontier models beat heuristic baselines and match the average human player but remain far behind experts. This gap widens as the run lengthens: the strongest models win the first act as often as experts, then fall behind. What separates strong models from weak is how well they play each fight; all of them misjudge risk over a run, preserving health rather than spending it to grow stronger. A comprehensive, self-consistent renaming of every game term shows that part of their performance rests on prior knowledge of the game, and models do not improve from the memories they write across runs, which are often wrong and overly cautious. Our harness is the first to …