{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/valse-a-task-independent-benchmark-for-vision-1","title":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena","arxiv_id":"2112.07566","date":"2021-12-14","proceeding":"ACL 2022 5","authors":["Letitia Parcalabescu","Michele Cafagna","Lilitta Muradjan","Anette Frank","Iacer Calixto","Albert Gatt"],"abstract":"We propose VALSE (Vision And Language Structured Evaluation), a novel benchmark designed for testing general-purpose pretrained vision and language (V&L) models for their visio-linguistic grounding capabilities on specific linguistic phenomena. VALSE offers a suite of six tests covering various linguistic constructs. Solving these requires models to ground linguistic phenomena in the visual modality, allowing more fine-grained evaluations than hitherto possible. We build VALSE using methods that support the construction of valid foils, and report results from evaluating five widely-used V&L models. Our experiments suggest that current models have considerable difficulty addressing most phenomena. Hence, we expect VALSE to serve as an important benchmark to measure future progress of pretrained V&L models from a linguistic perspective, complementing the canonical task-centred V&L evaluations.","url_abs":"https://arxiv.org/abs/2112.07566v2","url_pdf":"https://arxiv.org/pdf/2112.07566v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"valse-a-task-independent-benchmark-for-vision-1","repo_url":"https://github.com/heidelberg-nlp/valse","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"image-sentence-alignment","task_name":"image-sentence alignment"},{"task_slug":null,"task_name":"valid"}],"methods":[],"datasets_introduced":[{"slug":"valse","name":"VALSE","full_name":"VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena"}],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-sentence-alignment-on-valse","task":"image-sentence alignment","dataset":"VALSE","model":"ViLBERT 12-in-1","rank_in_archive_order":1,"of":7,"metrics":{"Average Accuracy":"63.2","average pairwise accuracy":"75.1"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse","task":"image-sentence alignment","dataset":"VALSE","model":"CLIP","rank_in_archive_order":2,"of":7,"metrics":{"average pairwise accuracy":"64.0"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse","task":"image-sentence alignment","dataset":"VALSE","model":"ViLBERT","rank_in_archive_order":3,"of":7,"metrics":{"Average Accuracy":"51.3","average pairwise accuracy":"63.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse","task":"image-sentence alignment","dataset":"VALSE","model":"GPT1","rank_in_archive_order":4,"of":7,"metrics":{"average pairwise accuracy":"60.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse","task":"image-sentence alignment","dataset":"VALSE","model":"GPT2","rank_in_archive_order":5,"of":7,"metrics":{"average pairwise accuracy":"60.1"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse","task":"image-sentence alignment","dataset":"VALSE","model":"LXMERT","rank_in_archive_order":6,"of":7,"metrics":{"Average Accuracy":"53.5","average pairwise accuracy":"59.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse","task":"image-sentence alignment","dataset":"VALSE","model":"VisualBERT","rank_in_archive_order":7,"of":7,"metrics":{"Average Accuracy":"48.8","average pairwise accuracy":"46.4"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-actant-swap","task":"image-sentence alignment","dataset":"VALSE actant swap","model":"GPT2","rank_in_archive_order":1,"of":7,"metrics":{"pairwise accuracy":"76.9"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-actant-swap","task":"image-sentence alignment","dataset":"VALSE actant swap","model":"GPT1","rank_in_archive_order":2,"of":7,"metrics":{"pairwise accuracy":"72.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-actant-swap","task":"image-sentence alignment","dataset":"VALSE actant swap","model":"CLIP","rank_in_archive_order":3,"of":7,"metrics":{"pairwise accuracy":"68.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-actant-swap","task":"image-sentence alignment","dataset":"VALSE actant swap","model":"ViLBERT","rank_in_archive_order":4,"of":7,"metrics":{"Accuracy (%)":"50.4","pairwise accuracy":"68.3"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-actant-swap","task":"image-sentence alignment","dataset":"VALSE actant swap","model":"ViLBERT 12-in-1","rank_in_archive_order":5,"of":7,"metrics":{"Accuracy (%)":"52.2","pairwise accuracy":"58.9"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-actant-swap","task":"image-sentence alignment","dataset":"VALSE actant swap","model":"LXMERT","rank_in_archive_order":6,"of":7,"metrics":{"Accuracy (%)":"48.5","pairwise accuracy":"45.8"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-actant-swap","task":"image-sentence alignment","dataset":"VALSE actant swap","model":"VisualBERT","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (%)":"49.7","pairwise accuracy":"44.4"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-action","task":"image-sentence alignment","dataset":"VALSE action replacement","model":"CLIP","rank_in_archive_order":1,"of":7,"metrics":{"pairwise accuracy":"75.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-action","task":"image-sentence alignment","dataset":"VALSE action replacement","model":"ViLBERT","rank_in_archive_order":2,"of":7,"metrics":{"Accuracy (%)":"52.6","pairwise accuracy":"70.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-action","task":"image-sentence alignment","dataset":"VALSE action replacement","model":"GPT2","rank_in_archive_order":3,"of":7,"metrics":{"pairwise accuracy":"66.8"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-action","task":"image-sentence alignment","dataset":"VALSE action replacement","model":"ViLBERT 12-in-1","rank_in_archive_order":4,"of":7,"metrics":{"Accuracy (%)":"57.3","pairwise accuracy":"65.9"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-action","task":"image-sentence alignment","dataset":"VALSE action replacement","model":"GPT1","rank_in_archive_order":5,"of":7,"metrics":{"pairwise accuracy":"65.4"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-action","task":"image-sentence alignment","dataset":"VALSE action replacement","model":"LXMERT","rank_in_archive_order":6,"of":7,"metrics":{"Accuracy (%)":"51.1","pairwise accuracy":"54.8"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-action","task":"image-sentence alignment","dataset":"VALSE action replacement","model":"VisualBERT","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (%)":"48.8","pairwise accuracy":"49.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference-1","task":"image-sentence alignment","dataset":"VALSE coreference clean","model":"ViLBERT 12-in-1","rank_in_archive_order":1,"of":7,"metrics":{"Accuracy (%)":"54.3","pairwise accuracy":"69.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference-1","task":"image-sentence alignment","dataset":"VALSE coreference clean","model":"GPT2","rank_in_archive_order":2,"of":7,"metrics":{"pairwise accuracy":"50.0"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference-1","task":"image-sentence alignment","dataset":"VALSE coreference clean","model":"CLIP","rank_in_archive_order":3,"of":7,"metrics":{"pairwise accuracy":"49.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference-1","task":"image-sentence alignment","dataset":"VALSE coreference clean","model":"ViLBERT","rank_in_archive_order":4,"of":7,"metrics":{"Accuracy (%)":"50.0","pairwise accuracy":"48.1"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference-1","task":"image-sentence alignment","dataset":"VALSE coreference clean","model":"VisualBERT","rank_in_archive_order":5,"of":7,"metrics":{"Accuracy (%)":"50.0","pairwise accuracy":"47.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference-1","task":"image-sentence alignment","dataset":"VALSE coreference clean","model":"GPT1","rank_in_archive_order":6,"of":7,"metrics":{"pairwise accuracy":"45.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference-1","task":"image-sentence alignment","dataset":"VALSE coreference clean","model":"LXMERT","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (%)":"49.0","pairwise accuracy":"44.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference","task":"image-sentence alignment","dataset":"VALSE coreference standard","model":"ViLBERT 12-in-1","rank_in_archive_order":1,"of":7,"metrics":{"Accuracy (%)":"54.4","pairwise accuracy":"75.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference","task":"image-sentence alignment","dataset":"VALSE coreference standard","model":"GPT2","rank_in_archive_order":2,"of":7,"metrics":{"pairwise accuracy":"54.5"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference","task":"image-sentence alignment","dataset":"VALSE coreference standard","model":"CLIP","rank_in_archive_order":3,"of":7,"metrics":{"pairwise accuracy":"52.1"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference","task":"image-sentence alignment","dataset":"VALSE coreference standard","model":"VisualBERT","rank_in_archive_order":4,"of":7,"metrics":{"Accuracy (%)":"50.0","pairwise accuracy":"49.5"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference","task":"image-sentence alignment","dataset":"VALSE coreference standard","model":"ViLBERT","rank_in_archive_order":5,"of":7,"metrics":{"Accuracy (%)":"50.0","pairwise accuracy":"47.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference","task":"image-sentence alignment","dataset":"VALSE coreference standard","model":"LXMERT","rank_in_archive_order":6,"of":7,"metrics":{"Accuracy (%)":"49.8","pairwise accuracy":"46.8"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-coreference","task":"image-sentence alignment","dataset":"VALSE coreference standard","model":"GPT1","rank_in_archive_order":7,"of":7,"metrics":{"pairwise accuracy":"45.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-2","task":"image-sentence alignment","dataset":"VALSE counting adversarial","model":"ViLBERT 12-in-1","rank_in_archive_order":1,"of":7,"metrics":{"Accuracy (%)":"66.7","pairwise accuracy":"77.3"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-2","task":"image-sentence alignment","dataset":"VALSE counting adversarial","model":"ViLBERT","rank_in_archive_order":2,"of":7,"metrics":{"Accuracy (%)":"51.8","pairwise accuracy":"73.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-2","task":"image-sentence alignment","dataset":"VALSE counting adversarial","model":"GPT1","rank_in_archive_order":3,"of":7,"metrics":{"pairwise accuracy":"69.5"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-2","task":"image-sentence alignment","dataset":"VALSE counting adversarial","model":"CLIP","rank_in_archive_order":4,"of":7,"metrics":{"pairwise accuracy":"57.5"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-2","task":"image-sentence alignment","dataset":"VALSE counting adversarial","model":"VisualBERT","rank_in_archive_order":5,"of":7,"metrics":{"Accuracy (%)":"50.0","pairwise accuracy":"50.0"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-2","task":"image-sentence alignment","dataset":"VALSE counting adversarial","model":"GPT2","rank_in_archive_order":6,"of":7,"metrics":{"pairwise accuracy":"45.3"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-2","task":"image-sentence alignment","dataset":"VALSE counting adversarial","model":"LXMERT","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (%)":"49.9","pairwise accuracy":"42.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting","task":"image-sentence alignment","dataset":"VALSE counting balanced","model":"ViLBERT 12-in-1","rank_in_archive_order":1,"of":7,"metrics":{"Accuracy (%)":"64.9","pairwise accuracy":"76.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting","task":"image-sentence alignment","dataset":"VALSE counting balanced","model":"LXMERT","rank_in_archive_order":2,"of":7,"metrics":{"Accuracy (%)":"52.0","pairwise accuracy":"62.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting","task":"image-sentence alignment","dataset":"VALSE counting balanced","model":"CLIP","rank_in_archive_order":3,"of":7,"metrics":{"pairwise accuracy":"62.1"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting","task":"image-sentence alignment","dataset":"VALSE counting balanced","model":"ViLBERT","rank_in_archive_order":4,"of":7,"metrics":{"Accuracy (%)":"50.7","pairwise accuracy":"58.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting","task":"image-sentence alignment","dataset":"VALSE counting balanced","model":"GPT2","rank_in_archive_order":5,"of":7,"metrics":{"pairwise accuracy":"51.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting","task":"image-sentence alignment","dataset":"VALSE counting balanced","model":"GPT1","rank_in_archive_order":6,"of":7,"metrics":{"pairwise accuracy":"51.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting","task":"image-sentence alignment","dataset":"VALSE counting balanced","model":"VisualBERT","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (%)":"48.3","pairwise accuracy":"48.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-1","task":"image-sentence alignment","dataset":"VALSE counting small numbers","model":"ViLBERT 12-in-1","rank_in_archive_order":1,"of":7,"metrics":{"Accuracy (%)":"69.2","pairwise accuracy":"80.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-1","task":"image-sentence alignment","dataset":"VALSE counting small numbers","model":"LXMERT","rank_in_archive_order":2,"of":7,"metrics":{"Accuracy (%)":"55.4","pairwise accuracy":"69.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-1","task":"image-sentence alignment","dataset":"VALSE counting small numbers","model":"ViLBERT","rank_in_archive_order":3,"of":7,"metrics":{"Accuracy (%)":"50.6","pairwise accuracy":"62.9"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-1","task":"image-sentence alignment","dataset":"VALSE counting small numbers","model":"CLIP","rank_in_archive_order":4,"of":7,"metrics":{"pairwise accuracy":"62.5"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-1","task":"image-sentence alignment","dataset":"VALSE counting small numbers","model":"GPT2","rank_in_archive_order":5,"of":7,"metrics":{"pairwise accuracy":"49.8"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-1","task":"image-sentence alignment","dataset":"VALSE counting small numbers","model":"GPT1","rank_in_archive_order":6,"of":7,"metrics":{"pairwise accuracy":"48.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-counting-1","task":"image-sentence alignment","dataset":"VALSE counting small numbers","model":"VisualBERT","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (%)":"47.8","pairwise accuracy":"48.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-existence","task":"image-sentence alignment","dataset":"VALSE existence","model":"ViLBERT 12-in-1","rank_in_archive_order":1,"of":7,"metrics":{"Accuracy (%)":"89.0","pairwise accuracy":"95.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-existence","task":"image-sentence alignment","dataset":"VALSE existence","model":"LXMERT","rank_in_archive_order":2,"of":7,"metrics":{"Accuracy (%)":"55.8","pairwise accuracy":"78.6"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-existence","task":"image-sentence alignment","dataset":"VALSE existence","model":"CLIP","rank_in_archive_order":3,"of":7,"metrics":{"pairwise accuracy":"66.9"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-existence","task":"image-sentence alignment","dataset":"VALSE existence","model":"ViLBERT","rank_in_archive_order":4,"of":7,"metrics":{"Accuracy (%)":"2.4","pairwise accuracy":"66.5"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-existence","task":"image-sentence alignment","dataset":"VALSE existence","model":"GPT1","rank_in_archive_order":5,"of":7,"metrics":{"pairwise accuracy":"61.8"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-existence","task":"image-sentence alignment","dataset":"VALSE existence","model":"GPT2","rank_in_archive_order":6,"of":7,"metrics":{"pairwise accuracy":"58.0"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-existence","task":"image-sentence alignment","dataset":"VALSE existence","model":"VisualBERT","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (%)":"49.3","pairwise accuracy":"39.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-foil-it","task":"image-sentence alignment","dataset":"VALSE foil-it (noun phrases)","model":"CLIP","rank_in_archive_order":1,"of":7,"metrics":{"pairwise accuracy":"88.8"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-foil-it","task":"image-sentence alignment","dataset":"VALSE foil-it (noun phrases)","model":"LXMERT","rank_in_archive_order":2,"of":7,"metrics":{"Accuracy (%)":"70.8","pairwise accuracy":"87.1"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-foil-it","task":"image-sentence alignment","dataset":"VALSE foil-it (noun phrases)","model":"ViLBERT 12-in-1","rank_in_archive_order":3,"of":7,"metrics":{"Accuracy (%)":"71.5","pairwise accuracy":"86.9"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-foil-it","task":"image-sentence alignment","dataset":"VALSE foil-it (noun phrases)","model":"ViLBERT","rank_in_archive_order":4,"of":7,"metrics":{"Accuracy (%)":"55.9","pairwise accuracy":"86.9"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-foil-it","task":"image-sentence alignment","dataset":"VALSE foil-it (noun phrases)","model":"GPT2","rank_in_archive_order":5,"of":7,"metrics":{"pairwise accuracy":"80.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-foil-it","task":"image-sentence alignment","dataset":"VALSE foil-it (noun phrases)","model":"GPT1","rank_in_archive_order":6,"of":7,"metrics":{"pairwise accuracy":"77.5"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-foil-it","task":"image-sentence alignment","dataset":"VALSE foil-it (noun phrases)","model":"VisualBERT","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (%)":"46.6","pairwise accuracy":"48.5"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-plurality","task":"image-sentence alignment","dataset":"VALSE plurality","model":"ViLBERT 12-in-1","rank_in_archive_order":1,"of":7,"metrics":{"Accuracy (%)":"62.0","pairwise accuracy":"72.4"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-plurality","task":"image-sentence alignment","dataset":"VALSE plurality","model":"LXMERT","rank_in_archive_order":2,"of":7,"metrics":{"Accuracy (%)":"55.1","pairwise accuracy":"64.4"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-plurality","task":"image-sentence alignment","dataset":"VALSE plurality","model":"ViLBERT","rank_in_archive_order":3,"of":7,"metrics":{"Accuracy (%)":"50.3","pairwise accuracy":"61.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-plurality","task":"image-sentence alignment","dataset":"VALSE plurality","model":"CLIP","rank_in_archive_order":4,"of":7,"metrics":{"pairwise accuracy":"56.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-plurality","task":"image-sentence alignment","dataset":"VALSE plurality","model":"GPT1","rank_in_archive_order":5,"of":7,"metrics":{"pairwise accuracy":"53.1"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-plurality","task":"image-sentence alignment","dataset":"VALSE plurality","model":"GPT2","rank_in_archive_order":6,"of":7,"metrics":{"pairwise accuracy":"51.9"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-plurality","task":"image-sentence alignment","dataset":"VALSE plurality","model":"VisualBERT","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (%)":"46.5","pairwise accuracy":"45.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-spatial","task":"image-sentence alignment","dataset":"VALSE spatial relations","model":"GPT1","rank_in_archive_order":1,"of":7,"metrics":{"pairwise accuracy":"77.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-spatial","task":"image-sentence alignment","dataset":"VALSE spatial relations","model":"GPT2","rank_in_archive_order":2,"of":7,"metrics":{"pairwise accuracy":"75.0"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-spatial","task":"image-sentence alignment","dataset":"VALSE spatial relations","model":"ViLBERT 12-in-1","rank_in_archive_order":3,"of":7,"metrics":{"Accuracy (%)":"53.4","pairwise accuracy":"67.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-spatial","task":"image-sentence alignment","dataset":"VALSE spatial relations","model":"CLIP","rank_in_archive_order":4,"of":7,"metrics":{"pairwise accuracy":"64.3"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-spatial","task":"image-sentence alignment","dataset":"VALSE spatial relations","model":"LXMERT","rank_in_archive_order":5,"of":7,"metrics":{"Accuracy (%)":"50.8","pairwise accuracy":"60.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-spatial","task":"image-sentence alignment","dataset":"VALSE spatial relations","model":"ViLBERT","rank_in_archive_order":6,"of":7,"metrics":{"Accuracy (%)":"49.9","pairwise accuracy":"57.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-sentence-alignment-on-valse-spatial","task":"image-sentence alignment","dataset":"VALSE spatial relations","model":"VisualBERT","rank_in_archive_order":7,"of":7,"metrics":{"Accuracy (%)":"49.3","pairwise accuracy":"39.7"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2112.07566","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2112.07566"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/heidelberg-nlp/valse","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":4,"unverified":5},"by_repo_kind":{"official":{"samples":9,"ran":4,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"e21bb85b8e1b4836","entry":"do_nms","repo":"heidelberg-nlp/valse","repo_kind":"official","path":"modeling_frcnn.py","file_url":"https://github.com/heidelberg-nlp/valse/blob/HEAD/modeling_frcnn.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e21bb85b8e1b4836"}},{"code_sha256_prefix":"899d1e2736236351","entry":"is_remote_url","repo":"heidelberg-nlp/valse","repo_kind":"official","path":"utils.py","file_url":"https://github.com/heidelberg-nlp/valse/blob/HEAD/utils.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"899d1e2736236351"}},{"code_sha256_prefix":"8a78a755bda274a5","entry":"load_labels","repo":"heidelberg-nlp/valse","repo_kind":"official","path":"utils.py","file_url":"https://github.com/heidelberg-nlp/valse/blob/HEAD/utils.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8a78a755bda274a5"}},{"code_sha256_prefix":"cfa4887da544c990","entry":"norm_box","repo":"heidelberg-nlp/valse","repo_kind":"official","path":"modeling_frcnn.py","file_url":"https://github.com/heidelberg-nlp/valse/blob/HEAD/modeling_frcnn.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"cfa4887da544c990"}},{"code_sha256_prefix":"3b772f0bbcecca2c","entry":"compute_ppl","repo":"heidelberg-nlp/valse","repo_kind":"official","path":"unimodal_valse_eval.py","file_url":"https://github.com/heidelberg-nlp/valse/blob/HEAD/unimodal_valse_eval.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"3b772f0bbcecca2c"}},{"code_sha256_prefix":"bd3a3f4bd8750da2","entry":"load_checkpoint","repo":"heidelberg-nlp/valse","repo_kind":"official","path":"utils.py","file_url":"https://github.com/heidelberg-nlp/valse/blob/HEAD/utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"bd3a3f4bd8750da2"}},{"code_sha256_prefix":"c06bd30b6fcc7498","entry":"pad_list_tensors","repo":"heidelberg-nlp/valse","repo_kind":"official","path":"modeling_frcnn.py","file_url":"https://github.com/heidelberg-nlp/valse/blob/HEAD/modeling_frcnn.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c06bd30b6fcc7498"}},{"code_sha256_prefix":"0ea1a5e4fb771a67","entry":"read_foil_dataset","repo":"heidelberg-nlp/valse","repo_kind":"official","path":"read_foil_dataset.py","file_url":"https://github.com/heidelberg-nlp/valse/blob/HEAD/read_foil_dataset.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0ea1a5e4fb771a67"}},{"code_sha256_prefix":"dc379b20add70eab","entry":"read_foils","repo":"heidelberg-nlp/valse","repo_kind":"official","path":"read_foil_dataset.py","file_url":"https://github.com/heidelberg-nlp/valse/blob/HEAD/read_foil_dataset.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"dc379b20add70eab"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}