{"url":"/sota/visual-question-answering-on-vqa-v2-test-std","task":{"name":"Visual Question Answering (VQA)","url":"/task/visual-question-answering","note":null},"dataset":{"name":"VQA v2 test-std","url":"/dataset/visual-question-answering-v2-0"},"category":"Computer Vision","categories":["Computer Vision","Natural Language Processing"],"category_note":null,"description":"**Visual Question Answering (VQA)** is a task in computer vision that involves answering questions about an image. The goal of VQA is to teach machines to understand the content of an image and answer questions about it in natural language.\r\n\r\nImage Source: [visualqa.org](https://visualqa.org/)","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["overall","yes/no","number","other"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"overall":null,"yes/no":null,"number":null,"other":null}},"counts":{"rows":38,"rows_with_code":34,"rows_with_paper_page":37,"rows_dated":37,"rows_using_additional_data":2},"rows":[{"rank_in_archive_order":1,"model":"BEiT-3","metrics":{"overall":"84.03"},"uses_additional_data":false,"paper_date":"2022-08-22","paper":"/paper/image-as-a-foreign-language-beit-pretraining","paper_url":"https://arxiv.org/abs/2208.10442v2","paper_title":"Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks","code":"https://github.com/microsoft/unilm/tree/master/beit","n_code_links":2,"syntology":null},{"rank_in_archive_order":2,"model":"mPLUG-Huge","metrics":{"number":"69.82","other":"77.02","overall":"83.62","yes/no":"94.83"},"uses_additional_data":false,"paper_date":"2022-05-24","paper":"/paper/mplug-effective-and-efficient-vision-language","paper_url":"https://arxiv.org/abs/2205.12005v2","paper_title":"mPLUG: Effective and Efficient Vision-Language Learning by Cross-modal Skip-connections","code":"https://github.com/modelscope/modelscope","n_code_links":3,"syntology":null},{"rank_in_archive_order":3,"model":"ONE-PEACE","metrics":{"number":"72.24","other":"74.15","overall":"82.52","yes/no":"94.85"},"uses_additional_data":false,"paper_date":"2023-05-18","paper":"/paper/one-peace-exploring-one-general","paper_url":"https://arxiv.org/abs/2305.11172v1","paper_title":"ONE-PEACE: Exploring One General Representation Model Toward Unlimited Modalities","code":"https://github.com/modelscope/modelscope","n_code_links":2,"syntology":{"n_ran":2,"n_unverified":5,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"X2-VLM (large)","metrics":{"overall":"81.8"},"uses_additional_data":false,"paper_date":"2022-11-22","paper":"/paper/x-2-vlm-all-in-one-pre-trained-model-for","paper_url":"https://arxiv.org/abs/2211.12402v2","paper_title":"X$^2$-VLM: All-In-One Pre-trained Model For Vision-Language Tasks","code":"https://github.com/zengyan-97/x-vlm","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":3,"n_samples":6,"n_pointer_only_licence":6}},{"rank_in_archive_order":5,"model":"VLMo","metrics":{"number":"67.26","other":"72.87","overall":"81.30","yes/no":"94.68"},"uses_additional_data":false,"paper_date":"2021-11-03","paper":"/paper/vlmo-unified-vision-language-pre-training","paper_url":"https://arxiv.org/abs/2111.02358v2","paper_title":"VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts","code":"https://github.com/microsoft/unilm/tree/master/vlmo","n_code_links":2,"syntology":null},{"rank_in_archive_order":6,"model":"SimVLM","metrics":{"overall":"80.34"},"uses_additional_data":false,"paper_date":"2021-08-24","paper":"/paper/simvlm-simple-visual-language-model","paper_url":"https://arxiv.org/abs/2108.10904v3","paper_title":"SimVLM: Simple Visual Language Model Pretraining with Weak Supervision","code":"https://github.com/yulong-XJTU/SimVLM","n_code_links":2,"syntology":{"n_ran":22,"n_unverified":15,"n_samples":37,"n_pointer_only_licence":28}},{"rank_in_archive_order":7,"model":"X2-VLM (base)","metrics":{"overall":"80.2"},"uses_additional_data":false,"paper_date":"2022-11-22","paper":"/paper/x-2-vlm-all-in-one-pre-trained-model-for","paper_url":"https://arxiv.org/abs/2211.12402v2","paper_title":"X$^2$-VLM: All-In-One Pre-trained Model For Vision-Language Tasks","code":"https://github.com/zengyan-97/x-vlm","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":3,"n_samples":6,"n_pointer_only_licence":6}},{"rank_in_archive_order":8,"model":"VAST","metrics":{"overall":"80.19"},"uses_additional_data":true,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":9,"model":"VALOR","metrics":{"overall":"78.62"},"uses_additional_data":true,"paper_date":"2023-04-17","paper":"/paper/valor-vision-audio-language-omni-perception","paper_url":"https://arxiv.org/abs/2304.08345v2","paper_title":"VALOR: Vision-Audio-Language Omni-Perception Pretraining Model and Dataset","code":"https://github.com/TXH-mercury/VALOR","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"Prompt Tuning","metrics":{"overall":"78.53"},"uses_additional_data":false,"paper_date":"2022-08-04","paper":"/paper/prompt-tuning-for-generative-multimodal","paper_url":"https://arxiv.org/abs/2208.02532v1","paper_title":"Prompt Tuning for Generative Multimodal Pretrained Models","code":"https://github.com/ofa-sys/ofa","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":11,"model":"Prismer","metrics":{"number":"61.39","other":"69.70","overall":"78.49","yes/no":"93.09"},"uses_additional_data":false,"paper_date":"2023-03-04","paper":"/paper/prismer-a-vision-language-model-with-an","paper_url":"https://arxiv.org/abs/2303.02506v3","paper_title":"Prismer: A Vision-Language Model with Multi-Task Experts","code":"https://github.com/nvlabs/prismer","n_code_links":2,"syntology":null},{"rank_in_archive_order":12,"model":"MSR + MS Cog. Svcs., X10 models","metrics":{"number":"62.55","other":"67.87","overall":"77.45","yes/no":"92.38"},"uses_additional_data":false,"paper_date":"2021-01-02","paper":"/paper/vinvl-making-visual-representations-matter-in","paper_url":"https://arxiv.org/abs/2101.00529v2","paper_title":"VinVL: Revisiting Visual Representations in Vision-Language Models","code":"https://github.com/microsoft/Oscar","n_code_links":7,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":1}},{"rank_in_archive_order":13,"model":"MSR + MS Cog. Svcs.","metrics":{"number":"61.5","other":"66.68","overall":"76.63","yes/no":"92.04"},"uses_additional_data":false,"paper_date":"2021-01-02","paper":"/paper/vinvl-making-visual-representations-matter-in","paper_url":"https://arxiv.org/abs/2101.00529v2","paper_title":"VinVL: Revisiting Visual Representations in Vision-Language Models","code":"https://github.com/microsoft/Oscar","n_code_links":7,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":1}},{"rank_in_archive_order":14,"model":"ALBEF (14M)","metrics":{"overall":"76.04"},"uses_additional_data":false,"paper_date":"2021-07-16","paper":"/paper/align-before-fuse-vision-and-language","paper_url":"https://arxiv.org/abs/2107.07651v2","paper_title":"Align before Fuse: Vision and Language Representation Learning with Momentum Distillation","code":"https://github.com/salesforce/lavis","n_code_links":6,"syntology":{"n_ran":4,"n_unverified":1,"n_samples":5,"n_pointer_only_licence":3}},{"rank_in_archive_order":15,"model":"BGN, ensemble","metrics":{"number":"61.13","other":"66.28","overall":"75.92","yes/no":"90.89"},"uses_additional_data":false,"paper_date":"2019-07-23","paper":"/paper/graph-reasoning-networks-for-visual-question","paper_url":"https://arxiv.org/abs/1907.09815v2","paper_title":"Bilinear Graph Networks for Visual Question Answering","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":16,"model":"ERNIE-ViL-single model","metrics":{"number":"56.79","other":"65.24","overall":"74.93","yes/no":"90.83"},"uses_additional_data":false,"paper_date":"2020-06-30","paper":"/paper/ernie-vil-knowledge-enhanced-vision-language","paper_url":"https://arxiv.org/abs/2006.16934v3","paper_title":"ERNIE-ViL: Knowledge Enhanced Vision-Language Representations Through Scene Graph","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":17,"model":"Single, w/o VLP","metrics":{"number":"58.01","other":"64.77","overall":"74.16","yes/no":"89.18"},"uses_additional_data":false,"paper_date":"2020-01-10","paper":"/paper/in-defense-of-grid-features-for-visual","paper_url":"https://arxiv.org/abs/2001.03615v2","paper_title":"In Defense of Grid Features for Visual Question Answering","code":"https://github.com/clip-vil/CLIP-ViL","n_code_links":2,"syntology":null},{"rank_in_archive_order":18,"model":"Single, w/o VLP","metrics":{"number":"58.62","other":"63.78","overall":"73.86","yes/no":"89.46"},"uses_additional_data":false,"paper_date":"2020-04-25","paper":"/paper/deep-multimodal-neural-architecture-search","paper_url":"https://arxiv.org/abs/2004.12070v2","paper_title":"Deep Multimodal Neural Architecture Search","code":"https://github.com/MILVLG/mmnas","n_code_links":1,"syntology":null},{"rank_in_archive_order":19,"model":"UNITER (Large)","metrics":{"overall":"73.4"},"uses_additional_data":false,"paper_date":"2019-09-25","paper":"/paper/uniter-learning-universal-image-text-1","paper_url":"https://arxiv.org/abs/1909.11740v3","paper_title":"UNITER: UNiversal Image-TExt Representation Learning","code":"https://github.com/ChenRocks/UNITER","n_code_links":7,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":2}},{"rank_in_archive_order":20,"model":"X-101 grid features + MCAN","metrics":{"overall":"72.71"},"uses_additional_data":false,"paper_date":"2020-01-10","paper":"/paper/in-defense-of-grid-features-for-visual","paper_url":"https://arxiv.org/abs/2001.03615v2","paper_title":"In Defense of Grid Features for Visual Question Answering","code":"https://github.com/clip-vil/CLIP-ViL","n_code_links":2,"syntology":null},{"rank_in_archive_order":21,"model":"LXMERT","metrics":{"overall":"72.5"},"uses_additional_data":false,"paper_date":"2019-08-20","paper":"/paper/lxmert-learning-cross-modality-encoder","paper_url":"https://arxiv.org/abs/1908.07490v3","paper_title":"LXMERT: Learning Cross-Modality Encoder Representations from Transformers","code":"https://github.com/huggingface/transformers","n_code_links":9,"syntology":{"n_ran":4,"n_unverified":11,"n_samples":15,"n_pointer_only_licence":3}},{"rank_in_archive_order":22,"model":"VL-BERTLARGE","metrics":{"overall":"72.2"},"uses_additional_data":false,"paper_date":"2019-08-22","paper":"/paper/vl-bert-pre-training-of-generic-visual","paper_url":"https://arxiv.org/abs/1908.08530v4","paper_title":"VL-BERT: Pre-training of Generic Visual-Linguistic Representations","code":"https://github.com/jackroos/VL-BERT","n_code_links":3,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":23,"model":"MCAN+VC","metrics":{"overall":"71.49"},"uses_additional_data":false,"paper_date":"2020-02-27","paper":"/paper/visual-commonsense-r-cnn","paper_url":"https://arxiv.org/abs/2002.12204v3","paper_title":"Visual Commonsense R-CNN","code":"https://github.com/Wangt-CN/VC-R-CNN","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":3,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":24,"model":"VisualBERT","metrics":{"overall":"71"},"uses_additional_data":false,"paper_date":"2019-08-09","paper":"/paper/visualbert-a-simple-and-performant-baseline","paper_url":"https://arxiv.org/abs/1908.03557v1","paper_title":"VisualBERT: A Simple and Performant Baseline for Vision and Language","code":"https://github.com/uclanlp/visualbert","n_code_links":10,"syntology":{"n_ran":4,"n_unverified":5,"n_samples":9,"n_pointer_only_licence":6}},{"rank_in_archive_order":25,"model":"MCANed-6","metrics":{"overall":"70.9"},"uses_additional_data":false,"paper_date":"2019-06-25","paper":"/paper/deep-modular-co-attention-networks-for-visual-1","paper_url":"https://arxiv.org/abs/1906.10770v1","paper_title":"Deep Modular Co-Attention Networks for Visual Question Answering","code":"https://github.com/MILVLG/mcan-vqa","n_code_links":7,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":26,"model":"Unified VLP","metrics":{"overall":"70.7"},"uses_additional_data":false,"paper_date":"2019-09-24","paper":"/paper/unified-vision-language-pre-training-for","paper_url":"https://arxiv.org/abs/1909.11059v3","paper_title":"Unified Vision-Language Pre-Training for Image Captioning and VQA","code":"https://github.com/rmokady/clip_prefix_caption","n_code_links":3,"syntology":{"n_ran":5,"n_unverified":9,"n_samples":14,"n_pointer_only_licence":13}},{"rank_in_archive_order":27,"model":"BAN+Glove+Counter","metrics":{"overall":"70.4"},"uses_additional_data":false,"paper_date":"2018-05-21","paper":"/paper/bilinear-attention-networks","paper_url":"http://arxiv.org/abs/1805.07932v2","paper_title":"Bilinear Attention Networks","code":"https://github.com/facebookresearch/mmf","n_code_links":8,"syntology":{"n_ran":4,"n_unverified":9,"n_samples":13,"n_pointer_only_licence":0}},{"rank_in_archive_order":28,"model":"Up-Down","metrics":{"overall":"70.34"},"uses_additional_data":false,"paper_date":"2017-07-25","paper":"/paper/bottom-up-and-top-down-attention-for-image","paper_url":"http://arxiv.org/abs/1707.07998v3","paper_title":"Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering","code":"https://github.com/facebookresearch/mmf","n_code_links":65,"syntology":{"n_ran":9,"n_unverified":0,"n_samples":9,"n_pointer_only_licence":6}},{"rank_in_archive_order":29,"model":"Image features from bottom-up attention (adaptive K, ensemble)","metrics":{"overall":"70.3"},"uses_additional_data":false,"paper_date":"2017-08-09","paper":"/paper/tips-and-tricks-for-visual-question-answering","paper_url":"http://arxiv.org/abs/1708.02711v1","paper_title":"Tips and Tricks for Visual Question Answering: Learnings from the 2017 Challenge","code":"https://github.com/peteanderson80/bottom-up-attention","n_code_links":10,"syntology":null},{"rank_in_archive_order":30,"model":"Caption VQA","metrics":{"overall":"69.7"},"uses_additional_data":false,"paper_date":"2019-06-03","paper":"/paper/190600513","paper_url":"https://arxiv.org/abs/1906.00513v3","paper_title":"Generating Question Relevant Captions to Aid Visual Question Answering","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":31,"model":"MuRel","metrics":{"overall":"68.4"},"uses_additional_data":false,"paper_date":"2019-02-25","paper":"/paper/murel-multimodal-relational-reasoning-for","paper_url":"http://arxiv.org/abs/1902.09487v1","paper_title":"MUREL: Multimodal Relational Reasoning for Visual Question Answering","code":"https://github.com/Cadene/murel.bootstrap.pytorch","n_code_links":1,"syntology":null},{"rank_in_archive_order":32,"model":"DMN","metrics":{"overall":"68.4"},"uses_additional_data":false,"paper_date":"2018-02-15","paper":"/paper/learning-to-count-objects-in-natural-images","paper_url":"http://arxiv.org/abs/1802.05766v1","paper_title":"Learning to Count Objects in Natural Images for Visual Question Answering","code":"https://github.com/Cyanogenoid/vqa-counting","n_code_links":1,"syntology":null},{"rank_in_archive_order":33,"model":"BLOCK","metrics":{"overall":"67.9"},"uses_additional_data":false,"paper_date":"2019-01-31","paper":"/paper/block-bilinear-superdiagonal-fusion-for","paper_url":"http://arxiv.org/abs/1902.00038v2","paper_title":"BLOCK: Bilinear Superdiagonal Fusion for Visual Question Answering and Visual Relationship Detection","code":"https://github.com/Cadene/block.bootstrap.pytorch","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":2,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":34,"model":"MUTAN","metrics":{"overall":"67.4"},"uses_additional_data":false,"paper_date":"2017-05-18","paper":"/paper/mutan-multimodal-tucker-fusion-for-visual","paper_url":"http://arxiv.org/abs/1705.06676v1","paper_title":"MUTAN: Multimodal Tucker Fusion for Visual Question Answering","code":"https://github.com/Cadene/vqa.pytorch","n_code_links":6,"syntology":null},{"rank_in_archive_order":35,"model":"2D continuous softmax","metrics":{"overall":"66.27"},"uses_additional_data":false,"paper_date":"2020-06-12","paper":"/paper/sparse-and-continuous-attention-mechanisms","paper_url":"https://arxiv.org/abs/2006.07214v3","paper_title":"Sparse and Continuous Attention Mechanisms","code":"https://github.com/deep-spin/mcan-vqa-continuous-attention","n_code_links":2,"syntology":{"n_ran":12,"n_unverified":3,"n_samples":15,"n_pointer_only_licence":2}},{"rank_in_archive_order":36,"model":"MCB [11, 12]","metrics":{"overall":"62.27"},"uses_additional_data":false,"paper_date":"2016-12-02","paper":"/paper/making-the-v-in-vqa-matter-elevating-the-role","paper_url":"http://arxiv.org/abs/1612.00837v3","paper_title":"Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering","code":"https://github.com/necla-ml/SNLI-VE","n_code_links":7,"syntology":null},{"rank_in_archive_order":37,"model":"Language-only","metrics":{"overall":"44.26"},"uses_additional_data":false,"paper_date":"2016-12-02","paper":"/paper/making-the-v-in-vqa-matter-elevating-the-role","paper_url":"http://arxiv.org/abs/1612.00837v3","paper_title":"Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering","code":"https://github.com/necla-ml/SNLI-VE","n_code_links":7,"syntology":null},{"rank_in_archive_order":38,"model":"Prior","metrics":{"overall":"25.98"},"uses_additional_data":false,"paper_date":"2016-12-02","paper":"/paper/making-the-v-in-vqa-matter-elevating-the-role","paper_url":"http://arxiv.org/abs/1612.00837v3","paper_title":"Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering","code":"https://github.com/necla-ml/SNLI-VE","n_code_links":7,"syntology":null}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,821 of the 9,623 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9623,"papers_checked":6821,"papers_extracted_not_yet_verified":65,"boards_without_verdict":29,"papers_not_yet_extracted":2737},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-25T09:33:49+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":19,"rows_with_any_sample_ran":16,"distinct_papers_with_graph_line":17,"distinct_papers_with_any_sample_ran":14,"samples_over_distinct_papers":{"n_ran":76,"n_unverified":68,"n_samples":144,"n_pointer_only_licence":71,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":81,"n_unverified":71,"n_samples":152,"n_pointer_only_licence":78,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}