{"task":"image-sentence alignment","dataset":"VALSE actant swap","metric_names":["pairwise accuracy","Accuracy (%)"],"rows":[{"id":107546,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"GPT2","metrics":{"pairwise accuracy":"76.9"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":107547,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"GPT1","metrics":{"pairwise accuracy":"72.2"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":107548,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"CLIP","metrics":{"pairwise accuracy":"68.6"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":107549,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"ViLBERT","metrics":{"Accuracy (%)":"50.4","pairwise accuracy":"68.3"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":107550,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"ViLBERT 12-in-1","metrics":{"Accuracy (%)":"52.2","pairwise accuracy":"58.9"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":107551,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"LXMERT","metrics":{"Accuracy (%)":"48.5","pairwise accuracy":"45.8"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":107552,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"VisualBERT","metrics":{"Accuracy (%)":"49.7","pairwise accuracy":"44.4"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":129470,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"GPT2","metrics":{"pairwise accuracy":"76.9"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":129471,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"GPT1","metrics":{"pairwise accuracy":"72.2"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":129472,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"CLIP","metrics":{"pairwise accuracy":"68.6"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":129473,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"ViLBERT","metrics":{"Accuracy (%)":"50.4","pairwise accuracy":"68.3"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":129474,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"ViLBERT 12-in-1","metrics":{"Accuracy (%)":"52.2","pairwise accuracy":"58.9"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":129475,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"LXMERT","metrics":{"Accuracy (%)":"48.5","pairwise accuracy":"45.8"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]},{"id":129476,"task":"image-sentence alignment","parent_task":"Multimodal Text and Image Classification","dataset":"VALSE actant swap","model_name":"VisualBERT","metrics":{"Accuracy (%)":"49.7","pairwise accuracy":"44.4"},"paper_url":"https://arxiv.org/abs/2112.07566v2","paper_title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","paper_date":"2021-12-14","code_links":[{"title":"heidelberg-nlp/valse","url":"https://github.com/heidelberg-nlp/valse"}],"metrics_order":"[\"pairwise accuracy\", \"Accuracy (%)\"]","area":"Natural Language Processing","uses_additional_data":0,"source":"archive","tags":[]}]}