paper-with-me

Papers

Unveiling Uncertainty: A Deep Dive into Calibration and Performance of Multimodal Large Language Models

2024-12-19 · Zijun Chen, WenBo Hu, Guande He, Zhijie Deng, Zheng Zhang, Richang Hong

Multimodal large language models (MLLMs) combine visual and textual data for tasks such as image captioning and visual question answering. Proper uncertainty calibration is crucial, yet challenging, for reliable use in areas like healthcare and autonomous driving. This paper investigates representative MLLMs, focusing on their calibration across various scenarios, including before and after visual fine-tuning, as well as before and after multimodal training of the base LLMs. We observed miscalibration in their performance, and at the same time, no significant differences in calibration across these scenarios. We also highlight how uncertainty differs between text and images and how their integration affects overall uncertainty. To better understand MLLMs' miscalibration and their ability to self-assess uncertainty, we construct the IDK (I don't know) dataset, which is key to evaluating how they handle unknowns. Our findings reveal that MLLMs tend to give answers rather than admit uncertainty, but this self-assessment improves with proper prompt adjustments. Finally, to calibrate MLLMs and enhance model reliability, we propose techniques such as temperature scaling and iterative prompt optimization. Our results provide insights into improving MLLMs for effective and responsible deployment in multimodal applications. Code and IDK dataset: https://github.com/hfutml/Calibration-MLLM.

📄 PDF Abstract BibTeX arXiv:2412.14660

Code (1)

hfutml/calibration-mllm 공식 구현 pytorch

Tasks

Autonomous DrivingImage CaptioningQuestion AnsweringVisual Question Answering

Methods 이 논문이 사용한 방법론

BASE 설명 없음

Similar Papers 제목 키워드 기반

On double-descent in uncertainty quantification in overparametrized models

2022-10-23 · Lucas Clarté, Bruno Loureiro, Florent Krzakala, Lenka Zdeborová

Uncertainty quantification is a central challenge in reliable and trustworthy machine learning. Naive measures such as last-layer scores are well-known to yield overconfident estimates in the context of overparametrized …

Binary ClassificationUncertainty Quantification

Mind the Confidence Gap: Overconfidence, Calibration, and Distractor Effects in Large Language Models

2025-02-16 · Prateek Chhikara

Large Language Models (LLMs) demonstrate impressive performance across diverse tasks, yet confidence calibration remains a challenge. Miscalibration - where models are overconfident or underconfident - poses risks, parti…

Multiple-choice

Combining Discrepancy-Confusion Uncertainty and Calibration Diversity for Active Fine-Grained Image Classification

2025-09-29 · Yinghao Jin, Xi Yang arxiv

Active learning (AL) aims to build high-quality labeled datasets by iteratively selecting the most informative samples from an unlabeled pool under limited annotation budgets. However, in fine-grained image classificatio…

Fine-Grained Image ClassificationActive Learning

Bi-level Meta-Policy Control for Dynamic Uncertainty Calibration in Evidential Deep Learning

2025-10-10 · Zhen Yang, Yansong Ma, Lei Chen arxiv

Traditional Evidence Deep Learning (EDL) methods rely on static hyperparameter for uncertainty calibration, limiting their adaptability in dynamic data distributions, which results in poor calibration and generalization …

Calibration Attention: Learning Reliability-Aware Representations for Vision Transformers

2025-08-12 · Wenhao Liang, Wei Emma Zhang, Lin Yue, Miao Xu 외 arxiv

Most calibration methods operate at the logit level, implicitly assuming that miscalibration can be corrected without changing the underlying representation. We challenge this assumption and propose \textbf{Calibration A…