提出融合模型直觉与反思的双通道置信度估计方法
Instinct vs. Reflection: Unifying Token and Verbalized Confidence in Multimodal Large Models
- 用隐式令牌支持与显式语言自评双信号融合置信度
- 在多模态大模型上提升置信度校准与错误预测能力
- 适合关注模型可靠性与安全部署的研究者
多模态大语言模型在感知与推理任务中表现优异，但实际应用中需可靠的置信度估计以保障可靠性。现有工作多聚焦于纯文本大模型，常依赖计算开销大的自一致性采样。本文将该问题拓展至多模态场景，对多模态大模型的响应置信度估计进行系统评估。分析发现模型隐式令牌级支持与其显式语言自评置信度存在显著偏差。为此，我们提出单调置信度融合框架，整合双通道信号并强化跨通道一致性，以估计判断正确性；随后引入保序均值对齐步骤，校正全局偏差，在保持选择性预测的风险-覆盖权衡前提下提升校准性能。在多种开源与闭源多模态大模型上的实验表明，本方法能持续生成更可靠的置信度估计，显著改善校准效果与失败预测能力。代码将发布于 https://github.com/Yunkaidang/Instinct-vs.-Reflection。
原文摘要 · Abstract (English)
Multimodal Large Language Models (MLLMs) have demonstrated exceptional capabilities in various perception and reasoning tasks. Despite this success, ensuring their reliability in practical deployment necessitates robust confidence estimation. Prior works have predominantly focused on text-only LLMs, often relying on computationally expensive self-consistency sampling. In this paper, we extend this to multimodal settings and conduct a comprehensive evaluation of MLLMs' response confidence estimation. Our analysis reveals a significant instinct-reflection misalignment: the model's implicit token-level support frequently diverges from its verbal self-assessment confidence. To address this misalignment, we propose a monotone confidence fusion framework to merge dual-channel signals and cross-channel consistency to estimate correctness. Subsequently, an order-preserving mean alignment step is applied to correct global bias, which improves calibration while preserving the risk-coverage trade-off for selective prediction. Experiments on diverse open-source and closed-source MLLMs show that our method consistently yields more reliable confidence estimates and improves both calibration and failure prediction. Code will be available at https://github.com/Yunkaidang/Instinct-vs.-Reflection.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。