发表机构
University of Science and Technology of China; Shanghai Jiao Tong University; Southwest Jiaotong University（中国科学技术大学; 上海交通大学; 西南交通大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究通过类型化快照-结算契约审计LLM智能体环境的动作结算，测试五种策略，发现保守拒绝完成率低，而优先级仲裁非最优，证据聚焦执行语义。
AI 中文摘要
在大语言模型（LLM）智能体环境中，并发动作即使每个提议单独有效，也需要仲裁。我们实现了一个类型化的快照-结算契约，并审计了三个不同的属性：顺序敏感性、有效进度和重放一致性。在28,800次穷举排列试验和2,160个脚本化多步情节中测试了五种结算策略。联合策略在固定优先级条件下具有空间顺序不变性，但保守拒绝在六智能体门口任务中仅完成31.25%的智能体，而随机票证完成率为90.28%；配对改进为59.03个百分点（95%自助法区间：50.00-68.06）。所有策略均保持所测试的空间约束，且优先级仲裁仍错过独立小实例最优解。一个单独的全状态日志审计精确重放了156个检查点，并拒绝了1,332个构造的损坏，且保留了终端锚点。证据涉及执行语义，而非人类真实性或长期公平性。
英文摘要
Concurrent actions in large language model (LLM) agent environments require arbitration even when each proposal is individually valid. We implement a typed snapshot-settlement contract and audit three distinct properties: order sensitivity, useful progress, and replay consistency. Five settlement policies are tested in 28,800 exhaustive permutation trials and 2,160 scripted multistep episodes. Joint policies are spatially order-invariant conditional on fixed priorities, yet conservative rejection completes only 31.25% of agents in a six-agent doorway task versus 90.28% for random tickets; the paired improvement is 59.03 percentage points (95% bootstrap interval: 50.00-68.06). All policies preserve the tested spatial constraints, and priority arbitration still misses the independent small-instance optimum. A separate full-state journal audit exactly replays 156 checkpoints and rejects 1,332 constructed corruptions with a retained terminal anchor. The evidence concerns execution semantics, not human realism or long-run fairness.
Comments5 pages, 2 figures
