arXiv：cs.LG· Hyeongwon Jang, Gyouk Chu, Changhun Kim, Hangyul Yoon, Jeonguk Lee, Eunho Yang, Joonhyung Park·· 2 天前AI 评分45
TRIAGE：面向不规则采样医疗时间序列可解释风险预测的辩证 LLM 推理框架
TRIAGE: Dialectical LLM Reasoning for Explainable Risk Prediction on Irregularly Sampled Medical Time Series
AI 导读
TRIAGE 通过让 LLM 对竞争性临床结果进行辩证推理，缓解了将分级风险压缩为过度自信预测的"风险极化"问题，使单一模型能同时输出临床依据和跨患者可比的 risk score。在五个 ISMTS 基准上，TRIAGE 相比竞争性 LLM 基线平均 AUPRC 提升 17.0%、校准误差降低 82.8%，并比最强 ISMTS 基线平均 AUPRC 高 3.5%。代码已开源。
正文
Abstract:Clinical early warning systems built on irregularly sampled medical time series (ISMTS) from electronic health records must deliver continuous risk scores for patient triage as well as interpretable rationales that clinicians can verify. Large language models (LLMs) are uniquely positioned for both, deriving risk from their output probabilities and rationales from their medical knowledge. However, we find that conventional LLM reasoning collapses graded risk into overconfident predictions and thereby undermines the cross-patient comparability on which triage depends. We refer to this failure mode as risk polarization and identify two underlying behaviors: early commitment to a single outcome, and one-sided reasoning that focuses only on the evidence for that outcome. To address this, we propose TRIAGE, a framework that trains an LLM to reason dialectically over competing clinical outcomes by eliciting outcome-specific rationales. This dialectical formulation mitigates risk polarization, enabling a single LLM to jointly provide explicit clinical rationales and risk scores comparable across patients. Across five ISMTS benchmarks, TRIAGE improves mean AUPRC by 17.0% and reduces mean calibration error by 82.8% relative to the competitive LLM-based baseline, while surpassing the strongest ISMTS baseline by 3.5% in mean AUPRC.
| Comments: | Code is available at this https URL |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2606.09030 [cs.LG] |
| (or arXiv:2606.09030v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2606.09030 arXiv-issued DOI via DataCite |
Submission history
From: Hyeongwon Jang [view email]
[v1]
Mon, 8 Jun 2026 04:53:44 UTC (500 KB)
[v2]
Thu, 1 Oct 2026 08:28:13 UTC (1,192 KB)
来源：arXiv：cs.LG · arxiv.org