多模态证据加贝叶斯主动问诊：合成评测中把家族诊断准确率从 87.0% 提到 95.5%
Auditable Hierarchical Inference with Multimodal Evidence and Bayesian Active Inquiry
中医多模态诊断的合成评测论文，把主动提问和客观证据分开测，连成本和失效边界都做了消融，做诊断类 Agent 的可以参考它的评测设计。
这篇 arXiv 论文构建了一个合成评测框架，用舌象、双侧多点脉搏和声学观测映射到 8 个证候家族，并用 25 个类型化问题对 33 个细项做贝叶斯式 narrowing。在预注册的 600 例合成评测中，仅客观推理的家族准确率为 87.0%，引入代价敏感期望信息增益（cost-EIG）后提升到 95.5%，细项准确率从 33.0% 提到 42.5%。但 cost-EIG 平均提问 9.795 次，多于固定顺序的 9.352 次，问题效率假设未通过；打乱细项答案似然会把家族准确率拉低到 78.7%。作者强调这是可复现的合成工作流，不构成临床诊断效力验证。
Auditable Hierarchical Inference with Multimodal Evidence and Bayesian Active Inquiry
Most multimodal classifiers assume that all useful evidence is available upfront, whereas clinical inquiry is sequential. We present a synthetic evaluation framework separating objective multimodal inference from active inquiry. An objective model maps synthetic tongue, bilateral multipressure pulse, and acoustic observations to eight syndrome families. A provenance-bearing Knowledge Core and training-only answer model support Bayesian narrowing over 33 fine terms using 25 typed questions, while seven safety questions remain outside the diagnostic posterior. Weak shared symptom factors limit fine-term influence on objective phenotypes, and an objective-only probe tests leakage. In a preregistered 600-case synthetic evaluation, objective family accuracy was 87.0%. Cost-sensitive expected information gain (cost-EIG) raised family accuracy to 95.5% and exact fine-term accuracy to 42.5%, versus 92.7% and 33.0% for an unoptimized fixed-order baseline. Selective family accuracy reached 96.9% at 95.3% coverage. However, cost-EIG asked more questions on average than fixed order (9.795 versus 9.352). Missing-audio routing retained 94.0% family accuracy, and a workflow invariant enforced abstention when all objective modalities were absent. Permuting fine-term answer likelihoods reduced family accuracy to 78.7%. Five of six frozen checks passed, but the question-efficiency hypothesis failed. Comparisons are descriptive because only aggregate, not paired case-level, outputs were retained. These results establish a reproducible synthetic workflow, not clinical diagnostic or treatment validity.