论文

Support-Gated World Model:用支持度门控提升重症升压药决策模拟的可靠性

When to Trust Simulated Patient Responses:Support-Gated World Models for Vasopressor Decision Support

精选理由

做医疗AI或临床智能体的开发者值得看:它量化了模拟预测何时该闭嘴。用支持度加集成分歧双重门控,MAP误差从9.2降到7.3 mmHg,代价是覆盖率从99.5%缩到81%。想学怎么给高风险场景的智能体加安全阀,这篇结构很清晰。

arXiv 论文提出 Support-Gated World-Model 决策支持框架,针对语言模型智能体在 propose-simulate-refine 流程中模拟不可靠的问题。系统由循环动力学集成预测短期生理反应,行为模型估计观察性动作支持度,两者共同决定是否输出预测,支持不足或集成分歧过大时直接扣留结果,无合格候选时选择弃权。

正文

在 eICU-CRD 数据集上覆盖 70 家医院、30,700 条决策转换,其中 14 家医院留作测试。62.0% 事实转换覆盖率下,联合门控的 MAP 平均绝对误差为 7.3 mmHg,随机选择为 9.2 mmHg;门控智能体在 81.0% 决策覆盖率下与记录动作一致率达 60.0%,未门控的集成智能体为 55.0% 一致率、99.5% 覆盖率。

论文明确指出结果是选择性临床模拟中可靠性-覆盖率权衡的刻画,并未识别个体化治疗效应或证明患者结局改善,属于医疗智能体安全使用模拟的边界研究。

原文 · arXiv cs.AI

When to Trust Simulated Patient Responses:Support-Gated World Models for Vasopressor Decision Support

Clinical world models expose language-model agents to predicted patient responses, but simulations can be unreliable for poorly supported treatments or unfamiliar hospitals. We introduce Support-Gated World-Model decision support, a lightweight extension of a propose--simulate--refine agent. A recurrent dynamics ensemble predicts short-horizon physiological responses to documented norepinephrine dose categories, while a behavior model estimates observational action support. The interface withholds predictions with insufficient support or excessive ensemble disagreement and abstains when no candidate qualifies. Evaluation on eICU-CRD includes 70 hospitals and 30,700 decision transitions, with 14 hospitals held out for testing. At 62.0% factual-transition coverage, joint gating yields a MAP mean absolute error of 7.3 mmHg, compared with 9.2 mmHg for random selection at the same coverage. The gated agent achieves 60.0% agreement with recorded actions at 81.0% decision coverage, compared with 55.0% agreement and 99.5% coverage for an ungated ensemble agent. Component ablations show lower factual risk with joint gating than with either gate alone.These results characterize a reliability--coverage trade-off in selective clinical simulation; they do not identify individualized treatment effects or establish improved patient outcomes.

产业坐标与信源核验

本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:world model、eICU-CRD、升压药决策、临床模拟,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。

信源核验说明:本文原始线索采编自「arXiv cs.AI」(@Yifan Jin, Tao Liu),经 AITOP 多维评分管线结构化提炼。