{"url":"/sota/zero-shot-video-question-answer-on-egoschema-1","task":{"name":"Zero-Shot Video Question Answer","url":"/task/zeroshot-video-question-answer","note":null},"dataset":{"name":"EgoSchema (fullset)","url":"/dataset/egoschema"},"category":"Computer Vision","categories":["Computer Vision","Miscellaneous","Natural Language Processing","Reasoning"],"category_note":null,"description":"This task present the results of Zeroshot Question Answer results on TGIF-QA dataset for LLM powered Video Conversational Models.","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":["Accuracy"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Accuracy":"higher"}},"counts":{"rows":29,"rows_with_code":28,"rows_with_paper_page":28,"rows_dated":28,"rows_using_additional_data":1},"rows":[{"rank_in_archive_order":1,"model":"BIMBA-LLaVA-Qwen2-7B","metrics":{"Accuracy":"71.14"},"uses_additional_data":false,"paper_date":"2025-03-12","paper":"/paper/bimba-selective-scan-compression-for-long","paper_url":"https://arxiv.org/abs/2503.09590v1","paper_title":"BIMBA: Selective-Scan Compression for Long-Range Video Question Answering","code":"https://github.com/md-mohaiminul/BIMBA","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"LinVT-Qwen2-VL(7B)","metrics":{"Accuracy":"69.5"},"uses_additional_data":false,"paper_date":"2024-12-06","paper":"/paper/linvt-empower-your-image-level-large-language","paper_url":"https://arxiv.org/abs/2412.05185v2","paper_title":"LinVT: Empower Your Image-level Large Language Model to Understand Videos","code":"https://github.com/gls0425/linvt","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":8,"n_samples":12,"n_pointer_only_licence":12}},{"rank_in_archive_order":3,"model":"Qwen2.5-Omni","metrics":{"Accuracy":"68.6"},"uses_additional_data":true,"paper_date":"2025-03-26","paper":"/paper/qwen2-5-omni-technical-report","paper_url":"https://arxiv.org/abs/2503.20215v1","paper_title":"Qwen2.5-Omni Technical Report","code":"https://github.com/QwenLM/Qwen2.5-Omni","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"LongVU (7B)","metrics":{"Accuracy":"67.6"},"uses_additional_data":false,"paper_date":"2024-10-22","paper":"/paper/longvu-spatiotemporal-adaptive-compression","paper_url":"https://arxiv.org/abs/2410.17434v1","paper_title":"LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding","code":"https://github.com/Vision-CAIR/LongVU","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":5,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"Video-RAG (Based on LLaVA-Video)","metrics":{"Accuracy":"66.7"},"uses_additional_data":false,"paper_date":"2024-11-20","paper":"/paper/video-rag-visually-aligned-retrieval","paper_url":"https://arxiv.org/abs/2411.13093v3","paper_title":"Video-RAG: Visually-aligned Retrieval-Augmented Long Video Comprehension","code":"https://github.com/leon1207/video-rag-master","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":6,"model":"VideoLLaMA2 (72B)","metrics":{"Accuracy":"63.9"},"uses_additional_data":false,"paper_date":"2024-06-11","paper":"/paper/videollama-2-advancing-spatial-temporal","paper_url":"https://arxiv.org/abs/2406.07476v3","paper_title":"VideoLLaMA 2: Advancing Spatial-Temporal Modeling and Audio Understanding in Video-LLMs","code":"https://github.com/damo-nlp-sg/videollama2","n_code_links":3,"syntology":{"n_ran":7,"n_unverified":10,"n_samples":17,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"Tarsier (34B)","metrics":{"Accuracy":"61.7"},"uses_additional_data":false,"paper_date":"2024-06-30","paper":"/paper/tarsier-recipes-for-training-and-evaluating-1","paper_url":"https://arxiv.org/abs/2407.00634v2","paper_title":"Tarsier: Recipes for Training and Evaluating Large Video Description Models","code":"https://github.com/bytedance/tarsier","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":8,"model":"LVNet","metrics":{"Accuracy":"61.1"},"uses_additional_data":false,"paper_date":"2024-06-13","paper":"/paper/too-many-frames-not-all-useful-efficient","paper_url":"https://arxiv.org/abs/2406.09396v4","paper_title":"Too Many Frames, Not All Useful: Efficient Strategies for Long-Form Video QA","code":"https://github.com/jongwoopark7978/LVNet","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":9,"model":"VideoTree (GPT4)","metrics":{"Accuracy":"61.1"},"uses_additional_data":false,"paper_date":"2024-05-29","paper":"/paper/videotree-adaptive-tree-based-video","paper_url":"https://arxiv.org/abs/2405.19209v2","paper_title":"VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos","code":"https://github.com/Ziyang412/VideoTree","n_code_links":1,"syntology":{"n_ran":11,"n_unverified":0,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":10,"model":"InternVideo2-6B","metrics":{"Accuracy":"60.2"},"uses_additional_data":false,"paper_date":"2024-03-22","paper":"/paper/internvideo2-scaling-video-foundation-models","paper_url":"https://arxiv.org/abs/2403.15377v4","paper_title":"InternVideo2: Scaling Foundation Models for Multimodal Video Understanding","code":"https://github.com/opengvlab/internvideo","n_code_links":2,"syntology":null},{"rank_in_archive_order":11,"model":"VideoChat-T (7B)","metrics":{"Accuracy":"60.0"},"uses_additional_data":false,"paper_date":"2024-10-25","paper":"/paper/timesuite-improving-mllms-for-long-video","paper_url":"https://arxiv.org/abs/2410.19702v2","paper_title":"TimeSuite: Improving MLLMs for Long Video Understanding via Grounded Tuning","code":"https://github.com/OpenGVLab/TimeSuite","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":3,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":12,"model":"VideoChat2_phi3","metrics":{"Accuracy":"56.7"},"uses_additional_data":false,"paper_date":"2023-11-28","paper":"/paper/mvbench-a-comprehensive-multi-modal-video","paper_url":"https://arxiv.org/abs/2311.17005v4","paper_title":"MVBench: A Comprehensive Multi-modal Video Understanding Benchmark","code":"https://github.com/opengvlab/ask-anything","n_code_links":3,"syntology":{"n_ran":7,"n_unverified":3,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":13,"model":"VideoChat2_HD_mistral","metrics":{"Accuracy":"55.8"},"uses_additional_data":false,"paper_date":"2023-11-28","paper":"/paper/mvbench-a-comprehensive-multi-modal-video","paper_url":"https://arxiv.org/abs/2311.17005v4","paper_title":"MVBench: A Comprehensive Multi-modal Video Understanding Benchmark","code":"https://github.com/opengvlab/ask-anything","n_code_links":3,"syntology":{"n_ran":7,"n_unverified":3,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":14,"model":"VideoChat2_mistral","metrics":{"Accuracy":"54.4"},"uses_additional_data":false,"paper_date":"2023-11-28","paper":"/paper/mvbench-a-comprehensive-multi-modal-video","paper_url":"https://arxiv.org/abs/2311.17005v4","paper_title":"MVBench: A Comprehensive Multi-modal Video Understanding Benchmark","code":"https://github.com/opengvlab/ask-anything","n_code_links":3,"syntology":{"n_ran":7,"n_unverified":3,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":15,"model":"Vamos (GPT-4o)","metrics":{"Accuracy":"53.6"},"uses_additional_data":false,"paper_date":"2023-11-22","paper":"/paper/vamos-versatile-action-models-for-video","paper_url":"https://arxiv.org/abs/2311.13627v3","paper_title":"Vamos: Versatile Action Models for Video Understanding","code":"https://github.com/brown-palm/Vamos","n_code_links":1,"syntology":{"n_ran":7,"n_unverified":1,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":16,"model":"TraveLER","metrics":{"Accuracy":"53.3"},"uses_additional_data":false,"paper_date":"2024-04-01","paper":"/paper/traveler-a-multi-lmm-agent-framework-for","paper_url":"https://arxiv.org/abs/2404.01476v2","paper_title":"TraveLER: A Modular Multi-LMM Agent Framework for Video Question-Answering","code":"https://github.com/traveler-framework/traveler","n_code_links":1,"syntology":null},{"rank_in_archive_order":17,"model":"LLoVi (GPT-3.5)","metrics":{"Accuracy":"50.3"},"uses_additional_data":false,"paper_date":"2023-12-28","paper":"/paper/a-simple-llm-framework-for-long-range-video","paper_url":"https://arxiv.org/abs/2312.17235v3","paper_title":"A Simple LLM Framework for Long-Range Video Question-Answering","code":"https://github.com/ceezh/llovi","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":0,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":18,"model":"Video ReCap","metrics":{"Accuracy":"50.23"},"uses_additional_data":false,"paper_date":"2024-02-20","paper":"/paper/video-recap-recursive-captioning-of-hour-long","paper_url":"https://arxiv.org/abs/2402.13250v6","paper_title":"Video ReCap: Recursive Captioning of Hour-Long Videos","code":"https://github.com/md-mohaiminul/VideoRecap","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":1}},{"rank_in_archive_order":19,"model":"Vamos (GPT-4)","metrics":{"Accuracy":"48.3"},"uses_additional_data":false,"paper_date":"2023-11-22","paper":"/paper/vamos-versatile-action-models-for-video","paper_url":"https://arxiv.org/abs/2311.13627v3","paper_title":"Vamos: Versatile Action Models for Video Understanding","code":"https://github.com/brown-palm/Vamos","n_code_links":1,"syntology":{"n_ran":7,"n_unverified":1,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":20,"model":"LangRepo (12B)","metrics":{"Accuracy":"41.2"},"uses_additional_data":false,"paper_date":"2024-03-21","paper":"/paper/language-repository-for-long-video","paper_url":"https://arxiv.org/abs/2403.14622v2","paper_title":"Language Repository for Long Video Understanding","code":"https://github.com/kkahatapitiya/langrepo","n_code_links":1,"syntology":{"n_ran":9,"n_unverified":0,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":21,"model":"MVU (13B)","metrics":{"Accuracy":"37.6"},"uses_additional_data":false,"paper_date":"2024-03-25","paper":"/paper/understanding-long-videos-in-one-multimodal","paper_url":"https://arxiv.org/abs/2403.16998v2","paper_title":"Understanding Long Videos with Multimodal Language Models","code":"https://github.com/kahnchana/mvu","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":0,"n_samples":4,"n_pointer_only_licence":1}},{"rank_in_archive_order":22,"model":"Vamos (13B)","metrics":{"Accuracy":"36.7"},"uses_additional_data":false,"paper_date":"2023-11-22","paper":"/paper/vamos-versatile-action-models-for-video","paper_url":"https://arxiv.org/abs/2311.13627v3","paper_title":"Vamos: Versatile Action Models for Video Understanding","code":"https://github.com/brown-palm/Vamos","n_code_links":1,"syntology":{"n_ran":7,"n_unverified":1,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":23,"model":"LLoVi (7B)","metrics":{"Accuracy":"33.5"},"uses_additional_data":false,"paper_date":"2023-12-28","paper":"/paper/a-simple-llm-framework-for-long-range-video","paper_url":"https://arxiv.org/abs/2312.17235v3","paper_title":"A Simple LLM Framework for Long-Range Video Question-Answering","code":"https://github.com/ceezh/llovi","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":0,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":24,"model":"TimeChat (7B)","metrics":{"Accuracy":"33.0"},"uses_additional_data":false,"paper_date":"2023-12-04","paper":"/paper/timechat-a-time-sensitive-multimodal-large","paper_url":"https://arxiv.org/abs/2312.02051v2","paper_title":"TimeChat: A Time-sensitive Multimodal Large Language Model for Long Video Understanding","code":"https://github.com/renshuhuai-andy/timechat","n_code_links":2,"syntology":{"n_ran":7,"n_unverified":4,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":25,"model":"InternVideo","metrics":{"Accuracy":"32.1"},"uses_additional_data":false,"paper_date":"2022-12-06","paper":"/paper/internvideo-general-video-foundation-models","paper_url":"https://arxiv.org/abs/2212.03191v2","paper_title":"InternVideo: General Video Foundation Models via Generative and Discriminative Learning","code":"https://github.com/opengvlab/internvideo","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":26,"model":"mPLUG-Owl (7B)","metrics":{"Accuracy":"31.1"},"uses_additional_data":false,"paper_date":"2023-04-27","paper":"/paper/mplug-owl-modularization-empowers-large","paper_url":"https://arxiv.org/abs/2304.14178v3","paper_title":"mPLUG-Owl: Modularization Empowers Large Language Models with Multimodality","code":"https://github.com/x-plug/mplug-owl","n_code_links":1,"syntology":null},{"rank_in_archive_order":27,"model":"FrozenBiLM","metrics":{"Accuracy":"26.9"},"uses_additional_data":false,"paper_date":"2022-06-16","paper":"/paper/zero-shot-video-question-answering-via-frozen","paper_url":"https://arxiv.org/abs/2206.08155v2","paper_title":"Zero-Shot Video Question Answering via Frozen Bidirectional Language Models","code":"https://github.com/antoyang/FrozenBiLM","n_code_links":3,"syntology":{"n_ran":14,"n_unverified":20,"n_samples":34,"n_pointer_only_licence":1}},{"rank_in_archive_order":28,"model":"SeViLA (4B)","metrics":{"Accuracy":"22.7"},"uses_additional_data":false,"paper_date":"2023-05-11","paper":"/paper/self-chained-image-language-model-for-video-1","paper_url":"https://arxiv.org/abs/2305.06988v2","paper_title":"Self-Chained Image-Language Model for Video Localization and Question Answering","code":"https://github.com/yui010206/sevila","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":1,"n_samples":7,"n_pointer_only_licence":7}},{"rank_in_archive_order":29,"model":"Random","metrics":{"Accuracy":"20.0"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"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 6,795 of the 9,581 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":9581,"papers_checked":6795,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2785},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+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":23,"rows_with_any_sample_ran":23,"distinct_papers_with_graph_line":18,"distinct_papers_with_any_sample_ran":18,"samples_over_distinct_papers":{"n_ran":103,"n_unverified":55,"n_samples":158,"n_pointer_only_licence":26,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":137,"n_unverified":63,"n_samples":200,"n_pointer_only_licence":26,"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"}}}