# Improving Automatic VQA Evaluation Using Large Language Models

arXiv: https://arxiv.org/abs/2310.02567 · HTML: https://mlsanity.com/paper/2310.02567/

## Key facts

- Cited by 37 papers in the map (data through 14 Oct 2026).
- Builds on 28 of its 38 references that are in the map.
- First posted on arXiv 4 Oct 2023.

## Builds on (most cited first, 28 of 28)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://mlsanity.com/paper/2201.12086/) (2022)
- [Finetuned Language Models Are Zero-Shot Learners](https://mlsanity.com/paper/2109.01652/) (2021)
- [OPT: Open Pre-trained Transformer Language Models](https://mlsanity.com/paper/2205.01068/) (2022)
- [Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations](https://mlsanity.com/paper/1602.07332/) (2016)
- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [VQA: Visual Question Answering](https://mlsanity.com/paper/1505.00468/) (2015)
- [G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment](https://mlsanity.com/paper/2303.16634/) (2023)
- [CIDEr: Consensus-based Image Description Evaluation](https://mlsanity.com/paper/1411.5726/) (2014)
- [OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge](https://mlsanity.com/paper/1906.00067/) (2019)
- [LIMA: Less Is More for Alignment](https://mlsanity.com/paper/2305.11206/) (2023)
- [On Faithfulness and Factuality in Abstractive Summarization](https://mlsanity.com/paper/2005.00661/) (2020)
- [Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks](https://mlsanity.com/paper/2204.07705/) (2022)
- [GPTScore: Evaluate as You Desire](https://mlsanity.com/paper/2302.04166/) (2023)
- [MIMIC-IT: Multi-Modal In-Context Instruction Tuning](https://mlsanity.com/paper/2306.05425/) (2023)
- [Evaluating Open-Domain Question Answering in the Era of Large Language Models](https://mlsanity.com/paper/2305.06984/) (2023)
- [PromptCap: Prompt-Guided Task-Aware Image Captioning](https://mlsanity.com/paper/2211.09699/) (2022)
- [Tomayto, Tomahto. Beyond Token-level Answer Equivalence for Question Answering Evaluation](https://mlsanity.com/paper/2202.07654/) (2022)
- [Semantic Answer Similarity for Evaluating Question Answering Models](https://mlsanity.com/paper/2108.06130/) (2021)
- [What's in a Name? Answer Equivalence For Open-Domain Question Answering](https://mlsanity.com/paper/2109.05289/) (2021)
- [Reassessing Evaluation Practices in Visual Question Answering: A Case Study on Out-of-Distribution Generalization](https://mlsanity.com/paper/2205.12191/) (2022)
- ['Just because you are right, doesn't mean I am wrong': Overcoming a Bottleneck in the Development and Evaluation of Open-Ended Visual Question Answering (VQA) Tasks](https://mlsanity.com/paper/2103.15022/) (2021)

## Built on by (most cited first, 30 of 37)

- [Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges](https://mlsanity.com/paper/2406.12624/) (2024)
- [Building and better understanding vision-language models: insights and future directions](https://mlsanity.com/paper/2408.12637/) (2024)
- [Prometheus-Vision: Vision-Language Model as a Judge for Fine-Grained Evaluation](https://mlsanity.com/paper/2401.06591/) (2024)
- [SPIQA: A Dataset for Multimodal Question Answering on Scientific Papers](https://mlsanity.com/paper/2407.09413/) (2024)
- [A Survey of State of the Art Large Vision Language Models: Alignment, Benchmark, Evaluations and Challenges](https://mlsanity.com/paper/2501.02189/) (2025)
- [Answer Matching Outperforms Multiple Choice for Language Model Evaluation](https://mlsanity.com/paper/2507.02856/) (2025)
- [Selective "Selective Prediction": Reducing Unnecessary Abstention in Vision-Language Reasoning](https://mlsanity.com/paper/2402.15610/) (2024)
- [Biomedical Visual Instruction Tuning with Clinician Preference Alignment](https://mlsanity.com/paper/2406.13173/) (2024)
- [Long-Form Answers to Visual Questions from Blind and Low Vision People](https://mlsanity.com/paper/2408.06303/) (2024)
- [Certainly Uncertain: A Benchmark and Metric for Multimodal Epistemic and Aleatoric Awareness](https://mlsanity.com/paper/2407.01942/) (2024)
- [Is Your Video Language Model a Reliable Judge?](https://mlsanity.com/paper/2503.05977/) (2025)
- [MSEarth: A Multimodal Benchmark for Earth Science Phenomenon Discovery with MLLMs](https://mlsanity.com/paper/2505.20740/) (2025)
- [MultiFinRAG: An Optimized Multimodal Retrieval-Augmented Generation (RAG) Framework for Financial Question Answering](https://mlsanity.com/paper/2506.20821/) (2025)
- [CLIP-UP: CLIP-Based Unanswerable Problem Detection for Visual Question Answering](https://mlsanity.com/paper/2501.01371/) (2025)
- [PairBench: Are Vision-Language Models Reliable at Comparing What They See?](https://mlsanity.com/paper/2502.15210/) (2025)
- [Generalizing Visual Question Answering from Synthetic to Human-Written Questions via a Chain of QA with a Large Language Model](https://mlsanity.com/paper/2401.06400/) (2024)
- [ERVQA: A Dataset to Benchmark the Readiness of Large Vision Language Models in Hospital Environments](https://mlsanity.com/paper/2410.06420/) (2024)
- [Describe Anything Model for Visual Question Answering on Text-rich Images](https://mlsanity.com/paper/2507.12441/) (2025)
- [EarthVL: A Progressive Earth Vision-Language Understanding and Generation Framework](https://mlsanity.com/paper/2601.02783/) (2026)
- [Beyond Accuracy: Benchmarking Cross-Task Consistency in Unified Multimodal Models](https://mlsanity.com/paper/2604.25072/) (2026)
- [UWBench: A Comprehensive Vision-Language Benchmark for Underwater Understanding](https://mlsanity.com/paper/2510.18262/) (2025)
- [HARMONY: Hidden Activation Representations and Model Output-Aware Uncertainty Estimation for Vision-Language Models](https://mlsanity.com/paper/2510.22171/) (2025)
- [Breaking Bad Molecules: Are MLLMs Ready for Structure-Level Molecular Detoxification?](https://mlsanity.com/paper/2506.10912/) (2025)
- [Judge, Retrieve, or Abstain: Uncertainty-Guarded LLM Judging with Provable Risk Guarantees](https://mlsanity.com/paper/2608.17994/) (2026)
- [SimpsonsVQA: Enhancing Inquiry-Based Learning with a Tailored Dataset](https://mlsanity.com/paper/2410.22648/) (2024)
- [All-in-one: Understanding and Generation in Multimodal Reasoning with the MAIA Benchmark](https://mlsanity.com/paper/2502.16989/) (2025)
- [Adapting Vision-Language Models for Evaluating World Models](https://mlsanity.com/paper/2506.17967/) (2025)
- [SMILE: A Composite Lexical-Semantic Metric for Question-Answering Evaluation](https://mlsanity.com/paper/2511.17432/) (2025)
- [Patent Figure Classification using Large Vision-language Models](https://mlsanity.com/paper/2501.12751/) (2025)
- [EmoAssist: Emotional Assistant for Visual Impairment Community](https://mlsanity.com/paper/2502.09285/) (2025)

## Abstract (arXiv)

8 years after the visual question answering (VQA) task was proposed, accuracy remains the primary metric for automatic evaluation. VQA Accuracy has been effective so far in the IID evaluation setting. However, our community is undergoing a shift towards open-ended generative models and OOD evaluation. In this new paradigm, the existing VQA Accuracy metric is overly stringent and underestimates the performance of VQA systems. Thus, there is a need to develop more robust automatic VQA metrics that serve as a proxy for human judgment. In this work, we propose to leverage the in-context learning capabilities of instruction-tuned large language models (LLMs) to build a better VQA metric. We formulate VQA evaluation as an answer-rating task where the LLM is instructed to score the accuracy of a candidate answer given a set of reference answers. We demonstrate the proposed metric better correlates with human judgment compared to existing metrics across several VQA models and benchmarks. We hope wide adoption of our metric will contribute to better estimating the research progress on the VQA task. We plan to release the evaluation code and collected human judgments.