摘要：大型语言模型（LLM）能够规划、使用工具、编写代码并执行长周期工作流，但强大的局部能力并不保证项目级执行控制。Agent 可能在原始目标已经达成后仍继续行动，产生低价值的细化、重复验证，以及对自身制造的复杂性进行修复。我们使用“LLM 帕金森主义”这一严格限定的非临床隐喻，描述这种尽管任务层面价值不断下降却仍持续行动的模式。我们认为，该问题不能仅由自回归式下一 Token 预测来解释；更直接的原因在于，动作提议生成、范围解释、进度评估和停止授权被集中于同一个自条件循环之中。因此，我们提出 Global Executive Control（GEC）v0.2：一种不确定性感知的治理架构，将动作生成与项目级控制分离。在一个包含 24,000 个回合、采用匹配候选项且统一设定 40,000 Token 上限的基准测试中，首候选基线的硬目标成功率为 67.42%，候选集局部控制的成功率为 96.53%，而 GEC 的成功率为 96.57%。候选集控制表明，获取多个候选动作解释了成功率提升中的大部分；相对于该控制方案，GEC 在保持成功率的同时，将平均 Token 使用量从 19,782 降至 12,574（下降 36.4%），并将 40,000 Token 上限下完成任务前的平均 Token 消耗从 16,136 降至 13,114（下降 18.7%）；同时消除了测得的完成前漂移，并显著降低总体复杂度。即使每个周期额外增加 500 个合成治理 Token，治理开销敏感性仍保持良好。这些机制模拟支持对范围、证据、资源使用和停止条件进行显式治理，但仍有必要进行真实模型验证。

Abstract:Large language models (LLMs) can plan, use tools, write code, and execute long-horizon workflows, yet strong local competence does not guarantee project-level executive control. Agents may continue acting after the original objective is satisfied, producing low-value refinements, repeated verification, and repairs to self-created complexity. We use LLM Parkinsonism as a narrowly defined, non-clinical metaphor for this pattern of persistent action despite diminishing task-level value. We argue that the problem is not explained by autoregressive next-token prediction alone, but more directly by concentrating proposal generation, scope interpretation, progress assessment, and stopping authority within the same self-conditioned loop. We therefore introduce Global Executive Control (GEC) v0.2, an uncertainty-aware governance architecture that separates action generation from project-level control. In a 24,000-episode matched-candidate benchmark under a common 40,000-token ceiling, a first-candidate baseline achieved 67.42% hard-goal success, a candidate-set local control achieved 96.53%, and GEC achieved 96.57%. The candidate-set control shows that access to multiple candidate actions explains most of the success gain; relative to that control, GEC preserved success while reducing mean token use from 19,782 to 12,574 (36.4%) and restricted mean tokens to completion at the 40,000-token ceiling from 16,136 to 13,114 (18.7%), while eliminating measured pre-completion drift and sharply reducing gross complexity. Governance-overhead sensitivity remained favorable through an additional 500 synthetic governance tokens per cycle. These mechanistic simulations support explicit governance of scope, evidence, resource use, and stopping, while live-model validation remains necessary.