MM-R$^3$: On (In-)Consistency of Vision-Language Models (VLMs)

• URL: http://arxiv.org/abs/2410.04778v2
• Date: Fri, 27 Jun 2025 17:08:56 GMT
• Title: MM-R$^3$: On (In-)Consistency of Vision-Language Models (VLMs)
• Authors: Shih-Han Chou, Shivam Chandhok, James J. Little, Leonid Sigal,
• Abstract summary: We analyze performance of SoTA Vision Language Models on three tasks: Question Rephrasing, Image Restyling, and Context Reasoning.<n>Our analysis reveals that consistency does not always align with accuracy, indicating that models with higher accuracy are not necessarily more consistent, and vice versa.<n>We propose a simple yet effective mitigation strategy in the form of an adapter module trained to minimize inconsistency across prompts.
• Score: 26.475993408532304
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: With the advent of LLMs and variants, a flurry of research has emerged, analyzing the performance of such models across an array of tasks. While most studies focus on evaluating the capabilities of state-of-the-art (SoTA) Vision Language Models (VLMs) through task accuracy (e.g., visual question answering, grounding), our work explores the related but complementary aspect of consistency - the ability of a VLM to produce semantically similar or identical responses to semantically similar queries. We note that consistency is a fundamental prerequisite (necessary but not sufficient condition) for robustness and trust in VLMs. Armed with this perspective, we propose the MM-R3 benchmark, which allows us to analyze performance, in terms of consistency and accuracy, of SoTA VLMs on three tasks: Question Rephrasing, Image Restyling, and Context Reasoning. Our analysis reveals that consistency does not always align with accuracy, indicating that models with higher accuracy are not necessarily more consistent, and vice versa. Furthermore, we propose a simple yet effective mitigation strategy in the form of an adapter module trained to minimize inconsistency across prompts. With our proposed strategy, we are able to achieve absolute improvements of 5.7% and 12.5%, on average on widely used VLMs such as BLIP-2 and LLaVa 1.5M in terms of consistency over their existing counterparts.

Related papers

• Test-Time Consistency in Vision Language Models [26.475993408532304]
  Vision-Language Models (VLMs) have achieved impressive performance across a wide range of multimodal tasks.<n>Recent benchmarks, such as MM-R3, highlight that even state-of-the-art VLMs can produce divergent predictions across semantically equivalent inputs.<n>We propose a simple and effective test-time consistency framework that enhances semantic consistency without supervised re-training.
  arXiv  Detail & Related papers  (2025-06-27T17:09:44Z)
• Consensus Entropy: Harnessing Multi-VLM Agreement for Self-Verifying and Self-Improving OCR [30.240680920617447]
  We introduce Consensus Entropy (CE), a training-free post-inference method that quantifies OCR uncertainty.<n>We develop a lightweight multi-model framework that effectively identifies problematic samples, selects the best outputs and combines model strengths.
  arXiv  Detail & Related papers  (2025-04-15T11:51:18Z)
• SCORE: Systematic COnsistency and Robustness Evaluation for Large Language Models [4.875712300661656]
  We present SCORE ($mathbfS$ystematic $mathbfCO$nsistency and $mathbfR$obustness $mathbfE$valuation), a comprehensive framework for non-adversarial evaluation of Large Language Models. The SCORE framework evaluates models by repeatedly testing them on the same benchmarks in various setups to give a realistic estimate of their accuracy and consistency.
  arXiv  Detail & Related papers  (2025-02-28T19:27:29Z)
• New Dataset and Methods for Fine-Grained Compositional Referring Expression Comprehension via Specialist-MLLM Collaboration [49.180693704510006]
  Referring Expression (REC) is a cross-modal task that evaluates the interplay of language understanding, image comprehension, and language-to-image grounding. We introduce a new REC dataset with two key features. First, it is designed with controllable difficulty levels, requiring fine-grained reasoning across object categories, attributes, and relationships. Second, it incorporates negative text and images generated through fine-grained editing, explicitly testing a model's ability to reject non-existent targets.
  arXiv  Detail & Related papers  (2025-02-27T13:58:44Z)
• VOILA: Evaluation of MLLMs For Perceptual Understanding and Analogical Reasoning [63.0285363282581]
  Multimodal Large Language Models (MLLMs) have become a powerful tool for integrating visual and textual information. We introduce VOILA, a benchmark designed to evaluate MLLMs' perceptual understanding and abstract relational reasoning. We reveal that current MLLMs struggle to comprehend inter-image relationships and exhibit limited capabilities in high-level relational reasoning.
  arXiv  Detail & Related papers  (2025-02-25T23:36:19Z)
• SPARC: Score Prompting and Adaptive Fusion for Zero-Shot Multi-Label Recognition in Vision-Language Models [74.40683913645731]
  Zero-shot multi-label recognition (MLR) with Vision-Language Models (VLMs) faces significant challenges without training data, model tuning, or architectural modifications.<n>Our work proposes a novel solution treating VLMs as black boxes, leveraging scores without training data or ground truth.<n>Analysis of these prompt scores reveals VLM biases and AND''/OR' signal ambiguities, notably that maximum scores are surprisingly suboptimal compared to second-highest scores.
  arXiv  Detail & Related papers  (2025-02-24T07:15:05Z)
• PairBench: Are Vision-Language Models Reliable at Comparing What They See? [16.49586486795478]
  We present PairBench, a framework to evaluate large vision language models (VLMs) for automatic evaluation depending on the task.<n>Our approach introduces four key metrics for reliable comparison: alignment with human annotations, consistency across pair ordering, distribution smoothness, and controllability through prompting.<n>Our analysis reveals that no model consistently excels across all metrics, with each demonstrating distinct strengths and weaknesses.
  arXiv  Detail & Related papers  (2025-02-21T04:53:11Z)
• DFPE: A Diverse Fingerprint Ensemble for Enhancing LLM Performance [11.753349115726952]
  We propose a novel ensemble method - Diverse Fingerprint Ensemble (DFPE) Our approach involves: (1) clustering models based on response "fingerprints" patterns, (2) applying a quantile-based filtering mechanism, and (3) assigning adaptive weights to remaining models. In experiments on the Massive Multitask Language Understanding (MMLU) benchmark, DFPE outperforms the best single model by 3% overall accuracy and 5% in discipline-level accuracy.
  arXiv  Detail & Related papers  (2025-01-29T08:44:45Z)
• AVTrustBench: Assessing and Enhancing Reliability and Robustness in Audio-Visual LLMs [70.4578433679737]
  We introduce Audio-Visual Trustworthiness assessment Benchmark (AVTrustBench), comprising 600K samples spanning over 9 meticulously crafted tasks.<n>Using our benchmark we extensively evaluate 13 state-of-the-art AVLLMs.<n>The findings reveal that the majority of existing models fall significantly short of achieving human-like comprehension.
  arXiv  Detail & Related papers  (2025-01-03T23:03:24Z)
• SURE-VQA: Systematic Understanding of Robustness Evaluation in Medical VQA Tasks [2.033441577169909]
  Vision-Language Models (VLMs) have great potential in medical tasks, like Visual Question Answering (VQA)<n>Their robustness to distribution shifts on unseen data remains a key concern for safe deployment.<n>We introduce a novel framework, called SURE-VQA, centered around three key requirements to overcome current pitfalls.
  arXiv  Detail & Related papers  (2024-11-29T13:22:52Z)
• VHELM: A Holistic Evaluation of Vision Language Models [75.88987277686914]
  We present the Holistic Evaluation of Vision Language Models (VHELM) VHELM aggregates various datasets to cover one or more of the 9 aspects: visual perception, knowledge, reasoning, bias, fairness, multilinguality, robustness, toxicity, and safety. Our framework is designed to be lightweight and automatic so that evaluation runs are cheap and fast.
  arXiv  Detail & Related papers  (2024-10-09T17:46:34Z)
• HELMET: How to Evaluate Long-Context Language Models Effectively and Thoroughly [34.205934899868346]
  We present HELMET, a comprehensive benchmark encompassing seven diverse, application-centric categories. We find that synthetic tasks like NIAH are not good predictors of downstream performance. While most LCLMs achieve perfect NIAH scores, open-source models significantly lag behind closed ones when the task requires full-context reasoning.
  arXiv  Detail & Related papers  (2024-10-03T17:20:11Z)
• SELF-GUIDE: Better Task-Specific Instruction Following via Self-Synthetic Finetuning [70.21358720599821]
  Large language models (LLMs) hold the promise of solving diverse tasks when provided with appropriate natural language prompts. We propose SELF-GUIDE, a multi-stage mechanism in which we synthesize task-specific input-output pairs from the student LLM. We report an absolute improvement of approximately 15% for classification tasks and 18% for generation tasks in the benchmark's metrics.
  arXiv  Detail & Related papers  (2024-07-16T04:41:58Z)
• MM-SpuBench: Towards Better Understanding of Spurious Biases in Multimodal LLMs [38.93090238335506]
  Spurious bias, a tendency to use spurious correlations between non-essential input attributes and target variables for predictions, has revealed a severe pitfall in deep learning models trained on single modality data. We introduce MM-SpuBench, a comprehensive visual question-answering (VQA) benchmark designed to evaluate MLLMs' reliance on nine distinct categories of spurious correlations. Our findings illuminate the persistence of the reliance on spurious correlations from these models and underscore the urge for new methodologies to mitigate spurious biases.
  arXiv  Detail & Related papers  (2024-06-24T20:29:16Z)
• MetaGPT: Merging Large Language Models Using Model Exclusive Task Arithmetic [6.46176287368784]
  We propose textbfModel textbfExclusive textbfTask textbfArithmetic for merging textbfGPT-scale models. Our proposed MetaGPT is data-agnostic and bypasses the heavy search process, making it cost-effective and easy to implement for LLMs.
  arXiv  Detail & Related papers  (2024-06-17T10:12:45Z)
• MMRel: A Relation Understanding Benchmark in the MLLM Era [72.95901753186227]
  Multi-Modal Relation Understanding (MMRel) is a benchmark that features large-scale, high-quality, and diverse data on inter-object relations. MMRel is ideal for evaluating MLLMs on relation understanding, as well as for fine-tuning MLLMs to enhance relation comprehension capability.
  arXiv  Detail & Related papers  (2024-06-13T13:51:59Z)
• Uncertainty Aware Learning for Language Model Alignment [97.36361196793929]
  We propose uncertainty-aware learning (UAL) to improve the model alignment of different task scenarios. We implement UAL in a simple fashion -- adaptively setting the label smoothing value of training according to the uncertainty of individual samples. Experiments on widely used benchmarks demonstrate that our UAL significantly and consistently outperforms standard supervised fine-tuning.
  arXiv  Detail & Related papers  (2024-06-07T11:37:45Z)
• MAML-en-LLM: Model Agnostic Meta-Training of LLMs for Improved In-Context Learning [43.512739869120125]
  We propose MAML-en-LLM, a novel method for meta-training large language models (LLMs) MAML-en-LLM can learn truly generalizable parameters that not only perform well on disjointed tasks but also adapts to unseen tasks. We demonstrate that MAML-en-LLM outperforms baselines in settings with limited amount of training data on both seen and unseen domains.
  arXiv  Detail & Related papers  (2024-05-19T04:49:42Z)
• PPTC-R benchmark: Towards Evaluating the Robustness of Large Language Models for PowerPoint Task Completion [96.47420221442397]
  We construct adversarial user instructions by attacking user instructions at sentence, semantic, and multi-language levels. We test 3 closed-source and 4 open-source LLMs using a benchmark that incorporates robustness settings. We find that GPT-4 exhibits the highest performance and strong robustness in our benchmark.
  arXiv  Detail & Related papers  (2024-03-06T15:33:32Z)
• Prismatic VLMs: Investigating the Design Space of Visually-Conditioned Language Models [73.40350756742231]
  Visually-conditioned language models (VLMs) have seen growing adoption in applications such as visual dialogue, scene understanding, and robotic task planning. Despite the volume of new releases, key design decisions around image preprocessing, architecture, and optimization are under-explored.
  arXiv  Detail & Related papers  (2024-02-12T18:21:14Z)
• MM-BigBench: Evaluating Multimodal Models on Multimodal Content Comprehension Tasks [56.60050181186531]
  We introduce MM-BigBench, which incorporates a diverse range of metrics to offer an extensive evaluation of the performance of various models and instructions. Our paper evaluates a total of 20 language models (14 MLLMs) on 14 multimodal datasets spanning 6 tasks, with 10 instructions for each task, and derives novel insights.
  arXiv  Detail & Related papers  (2023-10-13T11:57:04Z)
• Measuring and Improving Chain-of-Thought Reasoning in Vision-Language Models [61.28463542324576]
  Vision-language models (VLMs) have recently demonstrated strong efficacy as visual assistants that can generate human-like outputs. We evaluate existing state-of-the-art VLMs and find that even the best-performing model is unable to demonstrate strong visual reasoning capabilities and consistency. We propose a two-stage training framework aimed at improving both the reasoning performance and consistency of VLMs.
  arXiv  Detail & Related papers  (2023-09-08T17:49:44Z)
• Semantic Consistency for Assuring Reliability of Large Language Models [9.876355290198639]
  Large Language Models (LLMs) exhibit remarkable fluency and competence across various natural language tasks. We introduce a general measure of semantic consistency, and formulate multiple versions of this metric to evaluate the performance of various LLMs. We propose a novel prompting strategy, called Ask-to-Choose (A2C), to enhance semantic consistency.
  arXiv  Detail & Related papers  (2023-08-17T18:11:33Z)
• Two Failures of Self-Consistency in the Multi-Step Reasoning of LLMs [78.31625291513589]
  We argue that self-consistency is an important criteria for valid multi-step reasoning in tasks where the solution is composed of the answers to multiple sub-steps. We propose two types of self-consistency that are particularly important for multi-step reasoning -- hypothetical consistency and compositional consistency. We demonstrate that multiple variants of the GPT-3/-4 models exhibit poor consistency rates across both types of consistency on a variety of tasks.
  arXiv  Detail & Related papers  (2023-05-23T17:25:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.