{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:BFYZ65XAW2RTH3TBNOB5LKFDTB","short_pith_number":"pith:BFYZ65XA","schema_version":"1.0","canonical_sha256":"09719f76e0b6a333ee616b83d5a8a3984e7a0d09420d413aca2942464691d594","source":{"kind":"arxiv","id":"2104.08247","version":2},"attestation_state":"computed","paper":{"title":"What to Pre-Train on? Efficient Intermediate Task Selection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Andreas R\\\"uckl\\'e, Clifton Poth, Iryna Gurevych, Jonas Pfeiffer","submitted_at":"2021-04-16T17:31:18Z","abstract_excerpt":"Intermediate task fine-tuning has been shown to culminate in large transfer gains across many NLP tasks. With an abundance of candidate datasets as well as pre-trained language models, it has become infeasible to run the cross-product of all combinations to find the best transfer setting. In this work we first establish that similar sequential fine-tuning gains can be achieved in adapter settings, and subsequently consolidate previously proposed methods that efficiently identify beneficial tasks for intermediate transfer learning. We experiment with a diverse set of 42 intermediate and 11 targ"},"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":"2104.08247","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-04-16T17:31:18Z","cross_cats_sorted":[],"title_canon_sha256":"c68c6173be295b505e4e2955003c73948b959e7e9503254989f9a1c2dbee41e0","abstract_canon_sha256":"8657476a9f609e0494ec0db8ce3966786466039e1a476c17a727fd7f7b5b62c8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:13:03.039366Z","signature_b64":"PnN9hfxIu6GbAIgdeMGS/AbLGoztfKNbFL7gS7ZeJ4NnJTLSC6pCPZbR7u5Nngh1iD9yxy88OCgbvciutjLjDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"09719f76e0b6a333ee616b83d5a8a3984e7a0d09420d413aca2942464691d594","last_reissued_at":"2026-07-05T03:13:03.038923Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:13:03.038923Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"What to Pre-Train on? Efficient Intermediate Task Selection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Andreas R\\\"uckl\\'e, Clifton Poth, Iryna Gurevych, Jonas Pfeiffer","submitted_at":"2021-04-16T17:31:18Z","abstract_excerpt":"Intermediate task fine-tuning has been shown to culminate in large transfer gains across many NLP tasks. With an abundance of candidate datasets as well as pre-trained language models, it has become infeasible to run the cross-product of all combinations to find the best transfer setting. In this work we first establish that similar sequential fine-tuning gains can be achieved in adapter settings, and subsequently consolidate previously proposed methods that efficiently identify beneficial tasks for intermediate transfer learning. We experiment with a diverse set of 42 intermediate and 11 targ"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.08247","kind":"arxiv","version":2},"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/2104.08247/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":"2104.08247","created_at":"2026-07-05T03:13:03.038986+00:00"},{"alias_kind":"arxiv_version","alias_value":"2104.08247v2","created_at":"2026-07-05T03:13:03.038986+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.08247","created_at":"2026-07-05T03:13:03.038986+00:00"},{"alias_kind":"pith_short_12","alias_value":"BFYZ65XAW2RT","created_at":"2026-07-05T03:13:03.038986+00:00"},{"alias_kind":"pith_short_16","alias_value":"BFYZ65XAW2RTH3TB","created_at":"2026-07-05T03:13:03.038986+00:00"},{"alias_kind":"pith_short_8","alias_value":"BFYZ65XA","created_at":"2026-07-05T03:13:03.038986+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BFYZ65XAW2RTH3TBNOB5LKFDTB","json":"https://pith.science/pith/BFYZ65XAW2RTH3TBNOB5LKFDTB.json","graph_json":"https://pith.science/api/pith-number/BFYZ65XAW2RTH3TBNOB5LKFDTB/graph.json","events_json":"https://pith.science/api/pith-number/BFYZ65XAW2RTH3TBNOB5LKFDTB/events.json","paper":"https://pith.science/paper/BFYZ65XA"},"agent_actions":{"view_html":"https://pith.science/pith/BFYZ65XAW2RTH3TBNOB5LKFDTB","download_json":"https://pith.science/pith/BFYZ65XAW2RTH3TBNOB5LKFDTB.json","view_paper":"https://pith.science/paper/BFYZ65XA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2104.08247&json=true","fetch_graph":"https://pith.science/api/pith-number/BFYZ65XAW2RTH3TBNOB5LKFDTB/graph.json","fetch_events":"https://pith.science/api/pith-number/BFYZ65XAW2RTH3TBNOB5LKFDTB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BFYZ65XAW2RTH3TBNOB5LKFDTB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BFYZ65XAW2RTH3TBNOB5LKFDTB/action/storage_attestation","attest_author":"https://pith.science/pith/BFYZ65XAW2RTH3TBNOB5LKFDTB/action/author_attestation","sign_citation":"https://pith.science/pith/BFYZ65XAW2RTH3TBNOB5LKFDTB/action/citation_signature","submit_replication":"https://pith.science/pith/BFYZ65XAW2RTH3TBNOB5LKFDTB/action/replication_record"}},"created_at":"2026-07-05T03:13:03.038986+00:00","updated_at":"2026-07-05T03:13:03.038986+00:00"}