规避计算套利的计算困难性
原标题：Circumventing the Computational Hardness of Computational Arbitrage
AI 摘要
该论文将 AI 模型市场中的计算套利形式化为预算分配问题，并证明其一般情形为 NP-hard。作者发现，当推理支出在降低失败率上呈现对数边际收益递减时，目标函数为联合凹函数，可在大规模模型市场中高效求得全局最优解。研究还提出贪心查询排序规则和基于经验贝叶斯的样本高效效用估计方法，在多个智能体与编码基准上取得优于任何单一模型的成本-性能权衡，并给出边际效用均等、收入流向性价比高者、激励互补专业化等经济学结论。
正文节选
Circumventing the Computational Hardness of Computational Arbitrage Abstract AI model marketplaces host providers with heterogeneous costs and capabilities. These differences create an opportunity for computational arbitrage (Olmedo et al., 2026), in which an intermediary purchases inference from a portfolio of models and resells it to users at a profit. Beyond creating profit opportunities for the arbitrageur, computational arbitrage offers financial value to both consumers and providers and co