{"task":"Visual Reasoning","dataset":"Winoground","metric_names":["Text Score","Image Score","Group Score"],"rows":[{"id":31569,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"GPT-4o + CA","metrics":{"Group Score":"52","Image Score":"58.5","Text Score":"75.5"},"paper_url":"https://arxiv.org/abs/2501.13620v5","paper_title":"A Cognitive Paradigm Approach to Probe the Perception-Reasoning Interface in VLMs","paper_date":"2025-01-23","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31570,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"GPT-4V (CoT, pick b/w two options)","metrics":{"Group Score":"58.75","Image Score":"68.75","Text Score":"75.25"},"paper_url":"https://arxiv.org/abs/2311.09193v1","paper_title":"The Role of Chain-of-Thought in Complex Vision-Language Reasoning Task","paper_date":"2023-11-15","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31571,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"GPT-4V (pick b/w two options)","metrics":{"Group Score":"39.25","Image Score":"46.25","Text Score":"69.25"},"paper_url":"https://arxiv.org/abs/2311.09193v1","paper_title":"The Role of Chain-of-Thought in Complex Vision-Language Reasoning Task","paper_date":"2023-11-15","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31572,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"MMICL + CoCoT","metrics":{"Group Score":"50.75","Image Score":"52.5","Text Score":"64.25"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31573,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"GPT-4V + CoCoT","metrics":{"Group Score":"44.5","Image Score":"49.5","Text Score":"58.5"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31574,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OpenFlamingo + CoCoT","metrics":{"Group Score":"41.5","Image Score":"55.25","Text Score":"58.25"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31575,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"GPT-4V","metrics":{"Group Score":"37.75","Image Score":"42.5","Text Score":"54.5"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31576,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"FIBER (EqSim)","metrics":{"Group Score":"27.5","Image Score":"32.00","Text Score":"51.5"},"paper_url":"https://arxiv.org/abs/2303.14465v2","paper_title":"Equivariant Similarity for Vision-Language Foundation Models","paper_date":"2023-03-25","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31577,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"FIBER (finetuned, Flickr30k)","metrics":{"Group Score":"23.00","Image Score":"26.50","Text Score":"51.25"},"paper_url":"https://arxiv.org/abs/2303.14465v2","paper_title":"Equivariant Similarity for Vision-Language Foundation Models","paper_date":"2023-03-25","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31578,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"MMICL + CCoT","metrics":{"Group Score":"47.5","Image Score":"48","Text Score":"51"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31579,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OpenFlamingo + DDCoT","metrics":{"Group Score":"39","Image Score":"47.25","Text Score":"47.5"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31580,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"VQ2","metrics":{"Group Score":"30.5","Image Score":"42.2","Text Score":"47"},"paper_url":"https://arxiv.org/abs/2305.10400v4","paper_title":"What You See is What You Read? Improving Text-Image Alignment Evaluation","paper_date":"2023-05-17","code_links":[{"title":"yonatanbitton/wysiwyr","url":"https://github.com/yonatanbitton/wysiwyr"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31581,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"MMICL + DDCoT","metrics":{"Group Score":"36.75","Image Score":"45","Text Score":"46.75"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31582,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"X-VLM 16M","metrics":{"Group Score":"21.2","Image Score":"24.5","Text Score":"46.7"},"paper_url":"https://arxiv.org/abs/2305.07558v1","paper_title":"Measuring Progress in Fine-grained Vision-and-Language Understanding","paper_date":"2023-05-12","code_links":[{"title":"e-bug/fine-grained-evals","url":"https://github.com/e-bug/fine-grained-evals"},{"title":"e-bug/weak-relation-vlm","url":"https://github.com/e-bug/weak-relation-vlm"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31583,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"PaLI (ft SNLI-VE + Synthetic Data)","metrics":{"Group Score":"28.75","Image Score":"38","Text Score":"46.5"},"paper_url":"https://arxiv.org/abs/2305.10400v4","paper_title":"What You See is What You Read? Improving Text-Image Alignment Evaluation","paper_date":"2023-05-17","code_links":[{"title":"yonatanbitton/wysiwyr","url":"https://github.com/yonatanbitton/wysiwyr"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31584,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"FIBER","metrics":{"Group Score":"22.25","Image Score":"25.75","Text Score":"46.25"},"paper_url":"https://arxiv.org/abs/2303.14465v2","paper_title":"Equivariant Similarity for Vision-Language Foundation Models","paper_date":"2023-03-25","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31585,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"MMICL (FLAN-T5-XXL)","metrics":{"Group Score":"43.00","Image Score":"44.99","Text Score":"45.50"},"paper_url":"https://arxiv.org/abs/2309.07915v3","paper_title":"MMICL: Empowering Vision-language Model with Multi-Modal In-Context Learning","paper_date":"2023-09-14","code_links":[{"title":"haozhezhao/mic","url":"https://github.com/haozhezhao/mic"},{"title":"pkunlp-icler/mic","url":"https://github.com/pkunlp-icler/mic"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31586,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"PaLI (ft SNLI-VE)","metrics":{"Group Score":"28.70","Image Score":"41.50","Text Score":"45.00"},"paper_url":"https://arxiv.org/abs/2305.10400v4","paper_title":"What You See is What You Read? Improving Text-Image Alignment Evaluation","paper_date":"2023-05-17","code_links":[{"title":"yonatanbitton/wysiwyr","url":"https://github.com/yonatanbitton/wysiwyr"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31587,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"Gemini + DDCoT","metrics":{"Group Score":"23.75","Image Score":"25","Text Score":"45"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31588,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"METER (EqSim)","metrics":{"Group Score":"18.75","Image Score":"22.75","Text Score":"45.0"},"paper_url":"https://arxiv.org/abs/2303.14465v2","paper_title":"Equivariant Similarity for Vision-Language Foundation Models","paper_date":"2023-03-25","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31589,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"X-VLM 4M","metrics":{"Group Score":"21.5","Image Score":"26.7","Text Score":"44.0"},"paper_url":"https://arxiv.org/abs/2305.07558v1","paper_title":"Measuring Progress in Fine-grained Vision-and-Language Understanding","paper_date":"2023-05-12","code_links":[{"title":"e-bug/fine-grained-evals","url":"https://github.com/e-bug/fine-grained-evals"},{"title":"e-bug/weak-relation-vlm","url":"https://github.com/e-bug/weak-relation-vlm"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31590,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP2 (ft COCO)","metrics":{"Group Score":"23.50","Image Score":"26.00","Text Score":"44.00"},"paper_url":"https://arxiv.org/abs/2305.10400v4","paper_title":"What You See is What You Read? Improving Text-Image Alignment Evaluation","paper_date":"2023-05-17","code_links":[{"title":"yonatanbitton/wysiwyr","url":"https://github.com/yonatanbitton/wysiwyr"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31591,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"KeyComp* (GPT-4)","metrics":{"Group Score":"18.2","Image Score":"28.7","Text Score":"43.5"},"paper_url":"https://arxiv.org/abs/2401.11337v1","paper_title":"Prompting Large Vision-Language Models for Compositional Reasoning","paper_date":"2024-01-20","code_links":[{"title":"tossowski/keycomp","url":"https://github.com/tossowski/keycomp"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31592,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"METER (finetuned, Flickr30k)","metrics":{"Group Score":"14.75","Image Score":"20.75","Text Score":"43.5"},"paper_url":"https://arxiv.org/abs/2303.14465v2","paper_title":"Equivariant Similarity for Vision-Language Foundation Models","paper_date":"2023-03-25","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31593,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP2 (SGVL)","metrics":{"Group Score":"23.3","Image Score":"28.5","Text Score":"42.8"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31594,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP (SGVL)","metrics":{"Group Score":"21.5","Image Score":"27.3","Text Score":"42.8"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31595,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"KeyComp* (GPT-3.5)","metrics":{"Group Score":"17.4","Image Score":"27.8","Text Score":"42.7"},"paper_url":"https://arxiv.org/abs/2401.11337v1","paper_title":"Prompting Large Vision-Language Models for Compositional Reasoning","paper_date":"2024-01-20","code_links":[{"title":"tossowski/keycomp","url":"https://github.com/tossowski/keycomp"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31596,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OpenFlamingo + CCoT","metrics":{"Group Score":"20","Image Score":"27.5","Text Score":"42.5"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31597,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"NegBLIP","metrics":{"Group Score":"18.5","Image Score":"24.0","Text Score":"42.5"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31598,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"IAIS large (Flickr30k)","metrics":{"Group Score":"16.00","Image Score":"19.75","Text Score":"42.50"},"paper_url":"https://sslneurips22.github.io/paper_pdfs/paper_65.pdf","paper_title":"Does Structural Attention Improve Compositional Representations in Vision-Language Models?","paper_date":"2022-12-03","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31599,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"LLaVA-1.5-CCoT","metrics":{"Group Score":"22.3","Image Score":"35.5","Text Score":"42.0"},"paper_url":"https://arxiv.org/abs/2311.17076v3","paper_title":"Compositional Chain-of-Thought Prompting for Large Multimodal Models","paper_date":"2023-11-27","code_links":[{"title":"chancharikmitra/ccot","url":"https://github.com/chancharikmitra/ccot"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31600,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP2","metrics":{"Group Score":"19.0","Image Score":"23.8","Text Score":"42.0"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31601,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"IAIS large (COCO)","metrics":{"Group Score":"15.50","Image Score":"19.75","Text Score":"41.75"},"paper_url":"https://sslneurips22.github.io/paper_pdfs/paper_65.pdf","paper_title":"Does Structural Attention Improve Compositional Representations in Vision-Language Models?","paper_date":"2022-12-03","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31602,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"NegBLIP2","metrics":{"Group Score":"20.5","Image Score":"26.0","Text Score":"41.5"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31603,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP (+Graph Text, +Graph Neg)","metrics":{"Group Score":"19.0","Image Score":"25.5","Text Score":"40.5"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31604,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP (+Graph Text)","metrics":{"Group Score":"16.5","Image Score":"20.5","Text Score":"40.3"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31605,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"Gemini + CoCoT","metrics":{"Group Score":"27.75","Image Score":"32.5","Text Score":"40"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31606,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"CACR base","metrics":{"Group Score":"14.25","Image Score":"17.75","Text Score":"39.25"},"paper_url":"https://sslneurips22.github.io/paper_pdfs/paper_65.pdf","paper_title":"Does Structural Attention Improve Compositional Representations in Vision-Language Models?","paper_date":"2022-12-03","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31607,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"METER","metrics":{"Group Score":"12.00","Image Score":"15.75","Text Score":"39.25"},"paper_url":"https://arxiv.org/abs/2303.14465v2","paper_title":"Equivariant Similarity for Vision-Language Foundation Models","paper_date":"2023-03-25","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31608,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OpenFlamingo","metrics":{"Group Score":"33.25","Image Score":"41.25","Text Score":"39"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31609,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP","metrics":{"Group Score":"15.0","Image Score":"19.2","Text Score":"39.0"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31610,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"GPT-4V (image-caption match answer yes/no, zero-shot)","metrics":{"Group Score":"38.00","Image Score":"38.00","Text Score":"38.00"},"paper_url":null,"paper_title":null,"paper_date":null,"code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31611,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"UNITER large","metrics":{"Group Score":"10.50","Image Score":"14.00","Text Score":"38.00"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31612,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"VinVL","metrics":{"Group Score":"14.50","Image Score":"17.75","Text Score":"37.75"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31613,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"ViLLA large","metrics":{"Group Score":"11.00","Image Score":"13.25","Text Score":"37.00"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31614,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP (VisualGPTScore, α-tuned)","metrics":{"Group Score":"16.8","Image Score":"21.5","Text Score":"36.5"},"paper_url":"https://arxiv.org/abs/2306.01879v4","paper_title":"Revisiting the Role of Language Priors in Vision-Language Models","paper_date":"2023-06-02","code_links":[{"title":"linzhiqiu/visual_gpt_score","url":"https://github.com/linzhiqiu/visual_gpt_score"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31615,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP 14M","metrics":{"Group Score":"14.5","Image Score":"18.5","Text Score":"36.5"},"paper_url":"https://arxiv.org/abs/2305.07558v1","paper_title":"Measuring Progress in Fine-grained Vision-and-Language Understanding","paper_date":"2023-05-12","code_links":[{"title":"e-bug/fine-grained-evals","url":"https://github.com/e-bug/fine-grained-evals"},{"title":"e-bug/weak-relation-vlm","url":"https://github.com/e-bug/weak-relation-vlm"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31616,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"ViT-B/16 + BERT base + ViLEM","metrics":{"Text Score":"36.5"},"paper_url":"http://openaccess.thecvf.com//content/CVPR2023/html/Chen_ViLEM_Visual-Language_Error_Modeling_for_Image-Text_Retrieval_CVPR_2023_paper.html","paper_title":"ViLEM: Visual-Language Error Modeling for Image-Text Retrieval","paper_date":"2023-01-01","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31617,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"LLaVA-1.5","metrics":{"Group Score":"20.1","Image Score":"33.3","Text Score":"36.0"},"paper_url":"https://arxiv.org/abs/2311.17076v3","paper_title":"Compositional Chain-of-Thought Prompting for Large Multimodal Models","paper_date":"2023-11-27","code_links":[{"title":"chancharikmitra/ccot","url":"https://github.com/chancharikmitra/ccot"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31618,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP (ITM)","metrics":{"Group Score":"13.3","Image Score":"15.8","Text Score":"35.8"},"paper_url":"https://arxiv.org/abs/2306.01879v4","paper_title":"Revisiting the Role of Language Priors in Vision-Language Models","paper_date":"2023-06-02","code_links":[{"title":"linzhiqiu/visual_gpt_score","url":"https://github.com/linzhiqiu/visual_gpt_score"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31619,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP 129M","metrics":{"Group Score":"11.7","Image Score":"15.0","Text Score":"35.5"},"paper_url":"https://arxiv.org/abs/2305.07558v1","paper_title":"Measuring Progress in Fine-grained Vision-and-Language Understanding","paper_date":"2023-05-12","code_links":[{"title":"e-bug/fine-grained-evals","url":"https://github.com/e-bug/fine-grained-evals"},{"title":"e-bug/weak-relation-vlm","url":"https://github.com/e-bug/weak-relation-vlm"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31620,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"ROSITA (Flickr30k)","metrics":{"Group Score":"12.25","Image Score":"15.25","Text Score":"35.25"},"paper_url":"https://sslneurips22.github.io/paper_pdfs/paper_65.pdf","paper_title":"Does Structural Attention Improve Compositional Representations in Vision-Language Models?","paper_date":"2022-12-03","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31621,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"ViLT (ViT-B/32)","metrics":{"Group Score":"9.25","Image Score":"14.00","Text Score":"34.75"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31622,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP 129M (CapFilt/L)","metrics":{"Group Score":"12.2","Image Score":"15.2","Text Score":"34.7"},"paper_url":"https://arxiv.org/abs/2305.07558v1","paper_title":"Measuring Progress in Fine-grained Vision-and-Language Understanding","paper_date":"2023-05-12","code_links":[{"title":"e-bug/fine-grained-evals","url":"https://github.com/e-bug/fine-grained-evals"},{"title":"e-bug/weak-relation-vlm","url":"https://github.com/e-bug/weak-relation-vlm"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31623,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP-ViT/L 129M","metrics":{"Group Score":"12.2","Image Score":"14.5","Text Score":"34.7"},"paper_url":"https://arxiv.org/abs/2305.07558v1","paper_title":"Measuring Progress in Fine-grained Vision-and-Language Understanding","paper_date":"2023-05-12","code_links":[{"title":"e-bug/fine-grained-evals","url":"https://github.com/e-bug/fine-grained-evals"},{"title":"e-bug/weak-relation-vlm","url":"https://github.com/e-bug/weak-relation-vlm"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31624,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"Diffusion Classifier (zero-shot)","metrics":{"Text Score":"34.00"},"paper_url":"https://arxiv.org/abs/2303.16203v3","paper_title":"Your Diffusion Model is Secretly a Zero-Shot Classifier","paper_date":"2023-03-28","code_links":[{"title":"diffusion-classifier/diffusion-classifier","url":"https://github.com/diffusion-classifier/diffusion-classifier"},{"title":"SamsungSAILMontreal/ForestDiffusion","url":"https://github.com/SamsungSAILMontreal/ForestDiffusion"},{"title":"tajamul21/fate","url":"https://github.com/tajamul21/fate"},{"title":"LiYinqi/DIVE","url":"https://github.com/LiYinqi/DIVE"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31625,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"PEVL 14M","metrics":{"Group Score":"12.2","Image Score":"15.7","Text Score":"33.2"},"paper_url":"https://arxiv.org/abs/2305.07558v1","paper_title":"Measuring Progress in Fine-grained Vision-and-Language Understanding","paper_date":"2023-05-12","code_links":[{"title":"e-bug/fine-grained-evals","url":"https://github.com/e-bug/fine-grained-evals"},{"title":"e-bug/weak-relation-vlm","url":"https://github.com/e-bug/weak-relation-vlm"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31626,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"ALBEF 14M","metrics":{"Group Score":"12.7","Image Score":"16.2","Text Score":"32.5"},"paper_url":"https://arxiv.org/abs/2305.07558v1","paper_title":"Measuring Progress in Fine-grained Vision-and-Language Understanding","paper_date":"2023-05-12","code_links":[{"title":"e-bug/fine-grained-evals","url":"https://github.com/e-bug/fine-grained-evals"},{"title":"e-bug/weak-relation-vlm","url":"https://github.com/e-bug/weak-relation-vlm"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31627,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"FLAVA (ITM)","metrics":{"Group Score":"14.25","Image Score":"20.50","Text Score":"32.25"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31628,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"UNITER base","metrics":{"Group Score":"10.00","Image Score":"13.25","Text Score":"32.25"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31629,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"CLIP (SGVL)","metrics":{"Group Score":"9.8","Image Score":"14.0","Text Score":"32.0"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31630,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"ViT-B/16 + BERT base","metrics":{"Text Score":"31.2"},"paper_url":"http://openaccess.thecvf.com//content/CVPR2023/html/Chen_ViLEM_Visual-Language_Error_Modeling_for_Image-Text_Retrieval_CVPR_2023_paper.html","paper_title":"ViLEM: Visual-Language Error Modeling for Image-Text Retrieval","paper_date":"2023-01-01","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31631,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"Gemini","metrics":{"Group Score":"25","Image Score":"26","Text Score":"30.75"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31632,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OCLIP (ViT-H/14) ","metrics":{"Image Score":"12.75","Text Score":"30.75"},"paper_url":"https://arxiv.org/abs/2311.10708v2","paper_title":"SelfEval: Leveraging the discriminative nature of generative models for evaluation","paper_date":"2023-11-17","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31633,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"CLIP (ViT-B/32)","metrics":{"Group Score":"8.00","Image Score":"10.50","Text Score":"30.75"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31634,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OFA large (ITM)","metrics":{"Group Score":"7.25","Image Score":"10.25","Text Score":"30.75"},"paper_url":"https://arxiv.org/abs/2305.07021v1","paper_title":"Simple Token-Level Confidence Improves Caption Correctness","paper_date":"2023-05-11","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31635,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"KeyComp (GPT-3.5)","metrics":{"Group Score":"12.4","Image Score":"24.6","Text Score":"30.3"},"paper_url":"https://arxiv.org/abs/2401.11337v1","paper_title":"Prompting Large Vision-Language Models for Compositional Reasoning","paper_date":"2024-01-20","code_links":[{"title":"tossowski/keycomp","url":"https://github.com/tossowski/keycomp"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31636,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"CLIP (ViT-L/14)","metrics":{"Image Score":"8.0","Text Score":"30.25"},"paper_url":"https://arxiv.org/abs/2311.10708v2","paper_title":"SelfEval: Leveraging the discriminative nature of generative models for evaluation","paper_date":"2023-11-17","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31637,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"ViLLA base","metrics":{"Group Score":"8.00","Image Score":"12.00","Text Score":"30.00"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31638,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"syn-CLIP","metrics":{"Group Score":"9.50","Image Score":"11.50","Text Score":"30.00"},"paper_url":"https://arxiv.org/abs/2303.17590v2","paper_title":"Going Beyond Nouns With Vision & Language Models Using Synthetic Data","paper_date":"2023-03-30","code_links":[{"title":"uvavision/syvic","url":"https://github.com/uvavision/syvic"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31639,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"syn-CyCLIP","metrics":{"Group Score":"8.25","Image Score":"10.75","Text Score":"30.00"},"paper_url":"https://arxiv.org/abs/2303.17590v2","paper_title":"Going Beyond Nouns With Vision & Language Models Using Synthetic Data","paper_date":"2023-03-30","code_links":[{"title":"uvavision/syvic","url":"https://github.com/uvavision/syvic"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31640,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"NegCLIP","metrics":{"Group Score":"8.0","Image Score":"10.5","Text Score":"29.5"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31641,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OFA large (TLC-A)","metrics":{"Group Score":"17.50","Image Score":"27.00","Text Score":"29.25"},"paper_url":"https://arxiv.org/abs/2305.07021v1","paper_title":"Simple Token-Level Confidence Improves Caption Correctness","paper_date":"2023-05-11","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31642,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"ALBEF 4M","metrics":{"Group Score":"11.0","Image Score":"15.5","Text Score":"29.2"},"paper_url":"https://arxiv.org/abs/2305.07558v1","paper_title":"Measuring Progress in Fine-grained Vision-and-Language Understanding","paper_date":"2023-05-12","code_links":[{"title":"e-bug/fine-grained-evals","url":"https://github.com/e-bug/fine-grained-evals"},{"title":"e-bug/weak-relation-vlm","url":"https://github.com/e-bug/weak-relation-vlm"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31643,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"LDM-T5 (SelfEval)","metrics":{"Image Score":"13.50","Text Score":"29.00"},"paper_url":"https://arxiv.org/abs/2311.10708v2","paper_title":"SelfEval: Leveraging the discriminative nature of generative models for evaluation","paper_date":"2023-11-17","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31644,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"CyCLIP","metrics":{"Group Score":"7.25","Image Score":"9.50","Text Score":"28.50"},"paper_url":"https://arxiv.org/abs/2303.17590v2","paper_title":"Going Beyond Nouns With Vision & Language Models Using Synthetic Data","paper_date":"2023-03-30","code_links":[{"title":"uvavision/syvic","url":"https://github.com/uvavision/syvic"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31645,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"PDM-T5 (SelfEval)","metrics":{"Image Score":"12.00","Text Score":"28.25"},"paper_url":"https://arxiv.org/abs/2311.10708v2","paper_title":"SelfEval: Leveraging the discriminative nature of generative models for evaluation","paper_date":"2023-11-17","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31646,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"COCA ViT-L14 (f.t on COCO)","metrics":{"Group Score":"8.25","Image Score":"11.50","Text Score":"28.25"},"paper_url":"https://arxiv.org/abs/2305.10400v4","paper_title":"What You See is What You Read? Improving Text-Image Alignment Evaluation","paper_date":"2023-05-17","code_links":[{"title":"yonatanbitton/wysiwyr","url":"https://github.com/yonatanbitton/wysiwyr"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31647,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"LLaVA-1.5-ZS-CoT","metrics":{"Group Score":"12.3","Image Score":"22.5","Text Score":"28.0"},"paper_url":"https://arxiv.org/abs/2311.17076v3","paper_title":"Compositional Chain-of-Thought Prompting for Large Multimodal Models","paper_date":"2023-11-27","code_links":[{"title":"chancharikmitra/ccot","url":"https://github.com/chancharikmitra/ccot"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31648,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"BLIP (ITC)","metrics":{"Group Score":"6.5","Image Score":"9.0","Text Score":"28.0"},"paper_url":"https://arxiv.org/abs/2306.01879v4","paper_title":"Revisiting the Role of Language Priors in Vision-Language Models","paper_date":"2023-06-02","code_links":[{"title":"linzhiqiu/visual_gpt_score","url":"https://github.com/linzhiqiu/visual_gpt_score"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31649,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OFA large (ft SNLI-VE)","metrics":{"Group Score":"9.00","Image Score":"14.30","Text Score":"27.70"},"paper_url":"https://arxiv.org/abs/2305.10400v4","paper_title":"What You See is What You Read? Improving Text-Image Alignment Evaluation","paper_date":"2023-05-17","code_links":[{"title":"yonatanbitton/wysiwyr","url":"https://github.com/yonatanbitton/wysiwyr"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31650,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OFA base (ITM)","metrics":{"Group Score":"6.50","Image Score":"10.75","Text Score":"26.75"},"paper_url":"https://arxiv.org/abs/2305.07021v1","paper_title":"Simple Token-Level Confidence Improves Caption Correctness","paper_date":"2023-05-11","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31651,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"CLIP RN50x64","metrics":{"Group Score":"10.25","Image Score":"13.75","Text Score":"26.50"},"paper_url":"https://arxiv.org/abs/2305.10400v4","paper_title":"What You See is What You Read? Improving Text-Image Alignment Evaluation","paper_date":"2023-05-17","code_links":[{"title":"yonatanbitton/wysiwyr","url":"https://github.com/yonatanbitton/wysiwyr"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31652,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"LLaVA-7B (GPTScore)","metrics":{"Group Score":"10.50","Image Score":"17.00","Text Score":"25.50"},"paper_url":"https://arxiv.org/abs/2308.10509v2","paper_title":"An Examination of the Compositionality of Large Generative Vision-Language Models","paper_date":"2023-08-21","code_links":[{"title":"teleema/sade","url":"https://github.com/teleema/sade"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31653,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"FLAVA (contrastive)","metrics":{"Group Score":"9.00","Image Score":"13.50","Text Score":"25.25"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31654,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"Random chance","metrics":{"Group Score":"16.67","Image Score":"25.00","Text Score":"25.00"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31655,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"LLaVA","metrics":{"Group Score":"13.0","Image Score":"25.0","Text Score":"24.8"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31656,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OFA base (TLC-A)","metrics":{"Group Score":"13.75","Image Score":"23.50","Text Score":"24.50"},"paper_url":"https://arxiv.org/abs/2305.07021v1","paper_title":"Simple Token-Level Confidence Improves Caption Correctness","paper_date":"2023-05-11","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31657,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"MiniGPT-4-7B (GPTScore)","metrics":{"Group Score":"11.50","Image Score":"21.75","Text Score":"24.50"},"paper_url":"https://arxiv.org/abs/2308.10509v2","paper_title":"An Examination of the Compositionality of Large Generative Vision-Language Models","paper_date":"2023-08-21","code_links":[{"title":"teleema/sade","url":"https://github.com/teleema/sade"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31658,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"ViLBERT base","metrics":{"Group Score":"4.75","Image Score":"7.25","Text Score":"23.75"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31659,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"MiniGPT-4","metrics":{"Group Score":"9.5","Image Score":"18.0","Text Score":"23.3"},"paper_url":"https://arxiv.org/abs/2305.06343v2","paper_title":"Incorporating Structured Representations into Pretrained Vision & Language Models Using Scene Graphs","paper_date":"2023-05-10","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31660,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"MiniGPT-4-7B (VisualGPTScore)","metrics":{"Group Score":"9.50","Image Score":"18.00","Text Score":"23.25"},"paper_url":"https://arxiv.org/abs/2308.10509v2","paper_title":"An Examination of the Compositionality of Large Generative Vision-Language Models","paper_date":"2023-08-21","code_links":[{"title":"teleema/sade","url":"https://github.com/teleema/sade"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31661,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"VSE++ (COCO, ResNet)","metrics":{"Group Score":"4.00","Image Score":"8.00","Text Score":"22.75"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31662,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OFA tiny (ITM)","metrics":{"Group Score":"4.50","Image Score":"7.75","Text Score":"22.75"},"paper_url":"https://arxiv.org/abs/2305.07021v1","paper_title":"Simple Token-Level Confidence Improves Caption Correctness","paper_date":"2023-05-11","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31663,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"LDM-CLIP (SelfEval)","metrics":{"Image Score":"7.25","Text Score":"22.75"},"paper_url":"https://arxiv.org/abs/2311.10708v2","paper_title":"SelfEval: Leveraging the discriminative nature of generative models for evaluation","paper_date":"2023-11-17","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31664,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"Gemini + CCoT","metrics":{"Group Score":"20.75","Image Score":"33","Text Score":"22.5"},"paper_url":"https://arxiv.org/abs/2401.02582v1","paper_title":"CoCoT: Contrastive Chain-of-Thought Prompting for Large Multimodal Models with Multiple Image Inputs","paper_date":"2024-01-05","code_links":[{"title":"vista-h/gpt-4v_social_media","url":"https://github.com/vista-h/gpt-4v_social_media"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31665,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"InstructBLIP-CCoT ","metrics":{"Group Score":"8.3","Image Score":"21.3","Text Score":"21.0"},"paper_url":"https://arxiv.org/abs/2311.17076v3","paper_title":"Compositional Chain-of-Thought Prompting for Large Multimodal Models","paper_date":"2023-11-27","code_links":[{"title":"chancharikmitra/ccot","url":"https://github.com/chancharikmitra/ccot"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31666,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"VSRN (Flickr30k)","metrics":{"Group Score":"3.50","Image Score":"5.00","Text Score":"20.00"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31667,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"VSE++ (Flickr30k, ResNet)","metrics":{"Group Score":"2.75","Image Score":"5.00","Text Score":"20.00"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31668,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"VSE++ (Flickr30k, VGG)","metrics":{"Group Score":"4.50","Image Score":"6.25","Text Score":"19.75"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31669,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"UniT (ITM finetuned)","metrics":{"Group Score":"4.00","Image Score":"6.25","Text Score":"19.50"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31670,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"LXMERT","metrics":{"Group Score":"4.00","Image Score":"7.00","Text Score":"19.25"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31671,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"TIFA","metrics":{"Group Score":"11.30","Image Score":"12.50","Text Score":"19.00"},"paper_url":"https://arxiv.org/abs/2305.10400v4","paper_title":"What You See is What You Read? Improving Text-Image Alignment Evaluation","paper_date":"2023-05-17","code_links":[{"title":"yonatanbitton/wysiwyr","url":"https://github.com/yonatanbitton/wysiwyr"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31672,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"IDEFICS 80B","metrics":{"Group Score":"8.0","Image Score":"22.5","Text Score":"18.75"},"paper_url":null,"paper_title":null,"paper_date":null,"code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31673,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"VSE++ (COCO, VGG)","metrics":{"Group Score":"3.50","Image Score":"5.50","Text Score":"18.75"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31674,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"VSRN (COCO)","metrics":{"Group Score":"3.75","Image Score":"7.00","Text Score":"17.50"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31675,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"PDM-CLIP (SelfEval)","metrics":{"Image Score":"14.00","Text Score":"17.00"},"paper_url":"https://arxiv.org/abs/2311.10708v2","paper_title":"SelfEval: Leveraging the discriminative nature of generative models for evaluation","paper_date":"2023-11-17","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31676,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"IDEFICS 9B","metrics":{"Group Score":"5.0","Image Score":"20.8","Text Score":"16.8"},"paper_url":null,"paper_title":null,"paper_date":null,"code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31677,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"OFA tiny (TLC-A)","metrics":{"Group Score":"6.75","Image Score":"15.75","Text Score":"16.50"},"paper_url":"https://arxiv.org/abs/2305.07021v1","paper_title":"Simple Token-Level Confidence Improves Caption Correctness","paper_date":"2023-05-11","code_links":[],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31678,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"VisualBERT base","metrics":{"Group Score":"1.50","Image Score":"2.50","Text Score":"15.50"},"paper_url":"https://arxiv.org/abs/2204.03162v2","paper_title":"Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality","paper_date":"2022-04-07","code_links":[{"title":"wangt-cn/eqben","url":"https://github.com/wangt-cn/eqben"},{"title":"juletx/spatial-reasoning","url":"https://github.com/juletx/spatial-reasoning"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31679,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"MiniGPT-4-7B (BERTScore)","metrics":{"Group Score":"2.75","Image Score":"8.00","Text Score":"14.00"},"paper_url":"https://arxiv.org/abs/2308.10509v2","paper_title":"An Examination of the Compositionality of Large Generative Vision-Language Models","paper_date":"2023-08-21","code_links":[{"title":"teleema/sade","url":"https://github.com/teleema/sade"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31680,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"LLaVA-7B (BERTScore)","metrics":{"Group Score":"2.25","Image Score":"5.25","Text Score":"13.50"},"paper_url":"https://arxiv.org/abs/2308.10509v2","paper_title":"An Examination of the Compositionality of Large Generative Vision-Language Models","paper_date":"2023-08-21","code_links":[{"title":"teleema/sade","url":"https://github.com/teleema/sade"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31681,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"InstructBLIP-ZS-CoT","metrics":{"Group Score":"4.0","Image Score":"16.3","Text Score":"9.3"},"paper_url":"https://arxiv.org/abs/2311.17076v3","paper_title":"Compositional Chain-of-Thought Prompting for Large Multimodal Models","paper_date":"2023-11-27","code_links":[{"title":"chancharikmitra/ccot","url":"https://github.com/chancharikmitra/ccot"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]},{"id":31682,"task":"Visual Reasoning","parent_task":null,"dataset":"Winoground","model_name":"InstructBLIP","metrics":{"Group Score":"3.3","Image Score":"11.5","Text Score":"7.0"},"paper_url":"https://arxiv.org/abs/2311.17076v3","paper_title":"Compositional Chain-of-Thought Prompting for Large Multimodal Models","paper_date":"2023-11-27","code_links":[{"title":"chancharikmitra/ccot","url":"https://github.com/chancharikmitra/ccot"}],"metrics_order":"[\"Text Score\", \"Image Score\", \"Group Score\"]","area":"Reasoning","uses_additional_data":0,"source":"archive","tags":[]}]}