NAACL：針對RAG系統中大型語言模型的雜訊感知口語信心校準

摘要

在關鍵任務的事實領域中部署大型語言模型時，準確評估模型置信度至關重要。儘管檢索增強生成技術被廣泛應用於提升事實依據性，但RAG場景下的置信度校準機制仍未得到充分理解。我們在四個基準測試上展開系統性研究，發現由於檢索上下文存在噪聲，LLM表現出較差的校準性能。具體而言，矛盾或不相關的證據會加劇模型的虛假確定性，導致嚴重過度置信。為解決此問題，我們提出NAACL規則（噪聲感知置信校準規則），為噪聲下的過度置信問題建立理論基礎。基於這些規則，我們進一步設計NAACL框架，通過整合約2000個HotpotQA示例的監督信號，在無需依賴更強教師模型的情況下，藉助監督微調使模型具備內在的噪聲感知能力。實驗結果表明，NAACL帶來顯著提升，域內ECE分數改善10.9%，域外改善8.0%。通過橋接檢索噪聲與語言校準之間的鴻溝，NAACL為實現既精準又具認知可靠性的大型語言模型開闢了新路徑。

English

Accurately assessing model confidence is essential for deploying large language models (LLMs) in mission-critical factual domains. While retrieval-augmented generation (RAG) is widely adopted to improve grounding, confidence calibration in RAG settings remains poorly understood. We conduct a systematic study across four benchmarks, revealing that LLMs exhibit poor calibration performance due to noisy retrieved contexts. Specifically, contradictory or irrelevant evidence tends to inflate the model's false certainty, leading to severe overconfidence. To address this, we propose NAACL Rules (Noise-AwAre Confidence CaLibration Rules) to provide a principled foundation for resolving overconfidence under noise. We further design NAACL, a noise-aware calibration framework that synthesizes supervision from about 2K HotpotQA examples guided by these rules. By performing supervised fine-tuning (SFT) with this data, NAACL equips models with intrinsic noise awareness without relying on stronger teacher models. Empirical results show that NAACL yields substantial gains, improving ECE scores by 10.9% in-domain and 8.0% out-of-domain. By bridging the gap between retrieval noise and verbal calibration, NAACL paves the way for both accurate and epistemically reliable LLMs.

NAACL：針對RAG系統中大型語言模型的雜訊感知口語信心校準

NAACL: Noise-AwAre Verbal Confidence Calibration for LLMs in RAG Systems

摘要

Support