{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:QBVXMZ4T32WNTSWK5CH3HAIYUG","short_pith_number":"pith:QBVXMZ4T","schema_version":"1.0","canonical_sha256":"806b766793deacd9cacae88fb38118a1b1db2e2ef94200be16578cd14ca9bd87","source":{"kind":"arxiv","id":"2111.06719","version":2},"attestation_state":"computed","paper":{"title":"On Transferability of Prompt Tuning for Natural Language Processing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chi-Min Chan, Huadong Wang, Jie Zhou, Juanzi Li, Kaiyue Wen, Lei Hou, Maosong Sun, Peng Li, Xiaozhi Wang, Yankai Lin, Yujia Qin, Yusheng Su, Zhiyuan Liu","submitted_at":"2021-11-12T13:39:28Z","abstract_excerpt":"Prompt tuning (PT) is a promising parameter-efficient method to utilize extremely large pre-trained language models (PLMs), which can achieve comparable performance to full-parameter fine-tuning by only tuning a few soft prompts. However, PT requires much more training time than fine-tuning. Intuitively, knowledge transfer can help to improve the efficiency. To explore whether we can improve PT via prompt transfer, we empirically investigate the transferability of soft prompts across different downstream tasks and PLMs in this work. We find that (1) in zero-shot setting, trained soft prompts c"},"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":"2111.06719","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-11-12T13:39:28Z","cross_cats_sorted":[],"title_canon_sha256":"3171c8b4bf9e937ef893cc7655656d6ec9516c2a90b9a779a3baf2e2e369029d","abstract_canon_sha256":"fa83a28494427f8abe20bfe12091de16b9e50c3e2e9a6bf37d829cc69d5761c0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:24:44.942101Z","signature_b64":"2aUTWauh8tkBGJ/DGOmh3sJnaOKHiSKUFdP3i0VA7w43YiOvX5pzWcwaRHpEGktxftHasdB70WBKTiqyaSX8CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"806b766793deacd9cacae88fb38118a1b1db2e2ef94200be16578cd14ca9bd87","last_reissued_at":"2026-07-05T07:24:44.941546Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:24:44.941546Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On Transferability of Prompt Tuning for Natural Language Processing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chi-Min Chan, Huadong Wang, Jie Zhou, Juanzi Li, Kaiyue Wen, Lei Hou, Maosong Sun, Peng Li, Xiaozhi Wang, Yankai Lin, Yujia Qin, Yusheng Su, Zhiyuan Liu","submitted_at":"2021-11-12T13:39:28Z","abstract_excerpt":"Prompt tuning (PT) is a promising parameter-efficient method to utilize extremely large pre-trained language models (PLMs), which can achieve comparable performance to full-parameter fine-tuning by only tuning a few soft prompts. However, PT requires much more training time than fine-tuning. Intuitively, knowledge transfer can help to improve the efficiency. To explore whether we can improve PT via prompt transfer, we empirically investigate the transferability of soft prompts across different downstream tasks and PLMs in this work. We find that (1) in zero-shot setting, trained soft prompts c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2111.06719","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/2111.06719/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":"2111.06719","created_at":"2026-07-05T07:24:44.941604+00:00"},{"alias_kind":"arxiv_version","alias_value":"2111.06719v2","created_at":"2026-07-05T07:24:44.941604+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2111.06719","created_at":"2026-07-05T07:24:44.941604+00:00"},{"alias_kind":"pith_short_12","alias_value":"QBVXMZ4T32WN","created_at":"2026-07-05T07:24:44.941604+00:00"},{"alias_kind":"pith_short_16","alias_value":"QBVXMZ4T32WNTSWK","created_at":"2026-07-05T07:24:44.941604+00:00"},{"alias_kind":"pith_short_8","alias_value":"QBVXMZ4T","created_at":"2026-07-05T07:24:44.941604+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.14055","citing_title":"PEML: Parameter-efficient Multi-Task Learning with Optimized Continuous Prompts","ref_index":85,"is_internal_anchor":false},{"citing_arxiv_id":"2403.14608","citing_title":"Parameter-Efficient Fine-Tuning for Large Models: A Comprehensive Survey","ref_index":53,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QBVXMZ4T32WNTSWK5CH3HAIYUG","json":"https://pith.science/pith/QBVXMZ4T32WNTSWK5CH3HAIYUG.json","graph_json":"https://pith.science/api/pith-number/QBVXMZ4T32WNTSWK5CH3HAIYUG/graph.json","events_json":"https://pith.science/api/pith-number/QBVXMZ4T32WNTSWK5CH3HAIYUG/events.json","paper":"https://pith.science/paper/QBVXMZ4T"},"agent_actions":{"view_html":"https://pith.science/pith/QBVXMZ4T32WNTSWK5CH3HAIYUG","download_json":"https://pith.science/pith/QBVXMZ4T32WNTSWK5CH3HAIYUG.json","view_paper":"https://pith.science/paper/QBVXMZ4T","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2111.06719&json=true","fetch_graph":"https://pith.science/api/pith-number/QBVXMZ4T32WNTSWK5CH3HAIYUG/graph.json","fetch_events":"https://pith.science/api/pith-number/QBVXMZ4T32WNTSWK5CH3HAIYUG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QBVXMZ4T32WNTSWK5CH3HAIYUG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QBVXMZ4T32WNTSWK5CH3HAIYUG/action/storage_attestation","attest_author":"https://pith.science/pith/QBVXMZ4T32WNTSWK5CH3HAIYUG/action/author_attestation","sign_citation":"https://pith.science/pith/QBVXMZ4T32WNTSWK5CH3HAIYUG/action/citation_signature","submit_replication":"https://pith.science/pith/QBVXMZ4T32WNTSWK5CH3HAIYUG/action/replication_record"}},"created_at":"2026-07-05T07:24:44.941604+00:00","updated_at":"2026-07-05T07:24:44.941604+00:00"}