{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:XQYI46C4RTXD4AY5LT5IJLHVGF","short_pith_number":"pith:XQYI46C4","schema_version":"1.0","canonical_sha256":"bc308e785c8cee3e031d5cfa84acf53140a52ca434f5297976465527c7d48bc9","source":{"kind":"arxiv","id":"2107.13054","version":1},"attestation_state":"computed","paper":{"title":"Exceeding the Limits of Visual-Linguistic Multi-Task Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.CV","cs.LG"],"primary_cat":"cs.AI","authors_text":"Cameron R. Wolfe, Keld T. Lundgaard","submitted_at":"2021-07-27T19:42:14Z","abstract_excerpt":"By leveraging large amounts of product data collected across hundreds of live e-commerce websites, we construct 1000 unique classification tasks that share similarly-structured input data, comprised of both text and images. These classification tasks focus on learning the product hierarchy of different e-commerce websites, causing many of them to be correlated. Adopting a multi-modal transformer model, we solve these tasks in unison using multi-task learning (MTL). Extensive experiments are presented over an initial 100-task dataset to reveal best practices for \"large-scale MTL\" (i.e., MTL wit"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2107.13054","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2021-07-27T19:42:14Z","cross_cats_sorted":["cs.CL","cs.CV","cs.LG"],"title_canon_sha256":"80406174c9f23ea9f0879254b42ea146d30e92b259d912ea5f0c26f866c1a46c","abstract_canon_sha256":"c2f0de1453299030c2f28a1ced7e5ed310c45302d4c36c0e813fcb45a4232787"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:01:25.821777Z","signature_b64":"2uYF58JypfGj+SyTSFlWo8pckpAywC5Yvcg60wNSoAkppctBRgvzpmlCLNAMjCh/tDuOSN/C6eGaUYTTRDugAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bc308e785c8cee3e031d5cfa84acf53140a52ca434f5297976465527c7d48bc9","last_reissued_at":"2026-07-05T03:01:25.821250Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:01:25.821250Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Exceeding the Limits of Visual-Linguistic Multi-Task Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.CV","cs.LG"],"primary_cat":"cs.AI","authors_text":"Cameron R. Wolfe, Keld T. Lundgaard","submitted_at":"2021-07-27T19:42:14Z","abstract_excerpt":"By leveraging large amounts of product data collected across hundreds of live e-commerce websites, we construct 1000 unique classification tasks that share similarly-structured input data, comprised of both text and images. These classification tasks focus on learning the product hierarchy of different e-commerce websites, causing many of them to be correlated. Adopting a multi-modal transformer model, we solve these tasks in unison using multi-task learning (MTL). Extensive experiments are presented over an initial 100-task dataset to reveal best practices for \"large-scale MTL\" (i.e., MTL wit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2107.13054","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2107.13054/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2107.13054","created_at":"2026-07-05T03:01:25.821310+00:00"},{"alias_kind":"arxiv_version","alias_value":"2107.13054v1","created_at":"2026-07-05T03:01:25.821310+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2107.13054","created_at":"2026-07-05T03:01:25.821310+00:00"},{"alias_kind":"pith_short_12","alias_value":"XQYI46C4RTXD","created_at":"2026-07-05T03:01:25.821310+00:00"},{"alias_kind":"pith_short_16","alias_value":"XQYI46C4RTXD4AY5","created_at":"2026-07-05T03:01:25.821310+00:00"},{"alias_kind":"pith_short_8","alias_value":"XQYI46C4","created_at":"2026-07-05T03:01:25.821310+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2509.04982","citing_title":"Optimizing Small Transformer-Based Language Models for Multi-Label Sentiment Analysis in Short Texts","ref_index":20,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XQYI46C4RTXD4AY5LT5IJLHVGF","json":"https://pith.science/pith/XQYI46C4RTXD4AY5LT5IJLHVGF.json","graph_json":"https://pith.science/api/pith-number/XQYI46C4RTXD4AY5LT5IJLHVGF/graph.json","events_json":"https://pith.science/api/pith-number/XQYI46C4RTXD4AY5LT5IJLHVGF/events.json","paper":"https://pith.science/paper/XQYI46C4"},"agent_actions":{"view_html":"https://pith.science/pith/XQYI46C4RTXD4AY5LT5IJLHVGF","download_json":"https://pith.science/pith/XQYI46C4RTXD4AY5LT5IJLHVGF.json","view_paper":"https://pith.science/paper/XQYI46C4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2107.13054&json=true","fetch_graph":"https://pith.science/api/pith-number/XQYI46C4RTXD4AY5LT5IJLHVGF/graph.json","fetch_events":"https://pith.science/api/pith-number/XQYI46C4RTXD4AY5LT5IJLHVGF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XQYI46C4RTXD4AY5LT5IJLHVGF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XQYI46C4RTXD4AY5LT5IJLHVGF/action/storage_attestation","attest_author":"https://pith.science/pith/XQYI46C4RTXD4AY5LT5IJLHVGF/action/author_attestation","sign_citation":"https://pith.science/pith/XQYI46C4RTXD4AY5LT5IJLHVGF/action/citation_signature","submit_replication":"https://pith.science/pith/XQYI46C4RTXD4AY5LT5IJLHVGF/action/replication_record"}},"created_at":"2026-07-05T03:01:25.821310+00:00","updated_at":"2026-07-05T03:01:25.821310+00:00"}