---
id: 20260825-T0-03
title: "新方法抑制注意力偏见，提升AI在超长电子病历上的推理能力"
title_en: "Inhibitory Attention Reduces 'Lost-in-the-Middle' Errors in Long EHR Analysis"
url: https://ai.daily.yangsir.net/daily/20260825-T0-03
issue_date: 2026-08-25
publish_date: 2026-08-24T04:00:00.000Z
category: research
source_name: "arXiv cs.CL (NLP)"
source_url: https://arxiv.org/abs/2608.20348
---

# 新方法抑制注意力偏见，提升AI在超长电子病历上的推理能力

电子病历常超过10万token，大模型在处理时存在“lost-in-the-middle”现象，即对长文本中段信息的检索可靠度远低于首尾。研究者提出抑制性注意力机制来刻画和缓解这一问题。在临床长上下文推理任务中，该方法通过调整注意力分配，有效提升了对病历中段关键信息的识别准确率，为AI辅助诊断在真实复杂病历中的应用扫清了一个重要障碍。

## English Version

**Inhibitory Attention Reduces 'Lost-in-the-Middle' Errors in Long EHR Analysis**

This paper addresses the 'lost-in-the-middle' effect in LLMs when processing electronic health records (EHRs), which often exceed 100,000 tokens. The proposed inhibitory attention mechanism characterizes and mitigates this issue by recalibrating attention weights. In clinical reasoning tasks, it significantly improves the retrieval of critical information located in the middle of long contexts. This work tackles a key barrier to deploying AI for medical diagnosis on complex, real-world patient data.

---

**来源**：[arXiv cs.CL (NLP)](https://arxiv.org/abs/2608.20348)

**详情页**：https://ai.daily.yangsir.net/daily/20260825-T0-03

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*