{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:NNSEKU5WC6QH5THRC44EADVLJF","short_pith_number":"pith:NNSEKU5W","canonical_record":{"source":{"id":"1903.05987","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-03-14T13:32:31Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"ebf111e46ced68405a524bb3a5babf72a466b66b88e91660d52563ba3f650f17","abstract_canon_sha256":"4e209be446333f84609c163763e4312a1fe82b7d345bb2fbf2523d53fe7eb7cf"},"schema_version":"1.0"},"canonical_sha256":"6b644553b617a07eccf11738400eab49694329c8be0cf1cac0f7cde49532f444","source":{"kind":"arxiv","id":"1903.05987","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1903.05987","created_at":"2026-05-17T23:43:42Z"},{"alias_kind":"arxiv_version","alias_value":"1903.05987v2","created_at":"2026-05-17T23:43:42Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1903.05987","created_at":"2026-05-17T23:43:42Z"},{"alias_kind":"pith_short_12","alias_value":"NNSEKU5WC6QH","created_at":"2026-05-18T12:33:24Z"},{"alias_kind":"pith_short_16","alias_value":"NNSEKU5WC6QH5THR","created_at":"2026-05-18T12:33:24Z"},{"alias_kind":"pith_short_8","alias_value":"NNSEKU5W","created_at":"2026-05-18T12:33:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:NNSEKU5WC6QH5THRC44EADVLJF","target":"record","payload":{"canonical_record":{"source":{"id":"1903.05987","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-03-14T13:32:31Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"ebf111e46ced68405a524bb3a5babf72a466b66b88e91660d52563ba3f650f17","abstract_canon_sha256":"4e209be446333f84609c163763e4312a1fe82b7d345bb2fbf2523d53fe7eb7cf"},"schema_version":"1.0"},"canonical_sha256":"6b644553b617a07eccf11738400eab49694329c8be0cf1cac0f7cde49532f444","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:43:42.438252Z","signature_b64":"8fqx0ArUdgcw/Da9mWUGyq5nQzxggGKuHw+ceEcNNLAc5xgzalBcsbX4cPkh+z3v9lzaywQZ5VA1EO/8EMBqCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6b644553b617a07eccf11738400eab49694329c8be0cf1cac0f7cde49532f444","last_reissued_at":"2026-05-17T23:43:42.437617Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:43:42.437617Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1903.05987","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:43:42Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EtnRqh5xN8IDHykjQ+pyaIgaElMsJ/AhNBKQnjLnKO/+euUaCv3SsD6ayqZI3kz1Rz5ix+ZAuCnhEPQRyrihCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-29T14:37:00.016166Z"},"content_sha256":"0f875ad953a1720e6e63bdbbd9818343bb3b0a12aa702e4388bdecd0a45b975e","schema_version":"1.0","event_id":"sha256:0f875ad953a1720e6e63bdbbd9818343bb3b0a12aa702e4388bdecd0a45b975e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:NNSEKU5WC6QH5THRC44EADVLJF","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"To Tune or Not to Tune? Adapting Pretrained Representations to Diverse Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Matthew E. Peters, Noah A. Smith, Sebastian Ruder","submitted_at":"2019-03-14T13:32:31Z","abstract_excerpt":"While most previous work has focused on different pretraining objectives and architectures for transfer learning, we ask how to best adapt the pretrained model to a given target task. We focus on the two most common forms of adaptation, feature extraction (where the pretrained weights are frozen), and directly fine-tuning the pretrained model. Our empirical results across diverse NLP tasks with two state-of-the-art models show that the relative performance of fine-tuning vs. feature extraction depends on the similarity of the pretraining and target tasks. We explore possible explanations for t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1903.05987","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":""},"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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:43:42Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"q41qfMucS35WewgxR4mG3bNPTUu176iLhKddAOEjnCBJ8bkN4qKIXBnLyZrEbJLB6EiNv+fS9woVQc0ekegjCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-29T14:37:00.016967Z"},"content_sha256":"5be1ca24fbf9555ab3e16fff9824d06ac6748674f2fc591fcd219260b6df6e87","schema_version":"1.0","event_id":"sha256:5be1ca24fbf9555ab3e16fff9824d06ac6748674f2fc591fcd219260b6df6e87"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NNSEKU5WC6QH5THRC44EADVLJF/bundle.json","state_url":"https://pith.science/pith/NNSEKU5WC6QH5THRC44EADVLJF/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NNSEKU5WC6QH5THRC44EADVLJF/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-29T14:37:00Z","links":{"resolver":"https://pith.science/pith/NNSEKU5WC6QH5THRC44EADVLJF","bundle":"https://pith.science/pith/NNSEKU5WC6QH5THRC44EADVLJF/bundle.json","state":"https://pith.science/pith/NNSEKU5WC6QH5THRC44EADVLJF/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NNSEKU5WC6QH5THRC44EADVLJF/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:NNSEKU5WC6QH5THRC44EADVLJF","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"4e209be446333f84609c163763e4312a1fe82b7d345bb2fbf2523d53fe7eb7cf","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-03-14T13:32:31Z","title_canon_sha256":"ebf111e46ced68405a524bb3a5babf72a466b66b88e91660d52563ba3f650f17"},"schema_version":"1.0","source":{"id":"1903.05987","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1903.05987","created_at":"2026-05-17T23:43:42Z"},{"alias_kind":"arxiv_version","alias_value":"1903.05987v2","created_at":"2026-05-17T23:43:42Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1903.05987","created_at":"2026-05-17T23:43:42Z"},{"alias_kind":"pith_short_12","alias_value":"NNSEKU5WC6QH","created_at":"2026-05-18T12:33:24Z"},{"alias_kind":"pith_short_16","alias_value":"NNSEKU5WC6QH5THR","created_at":"2026-05-18T12:33:24Z"},{"alias_kind":"pith_short_8","alias_value":"NNSEKU5W","created_at":"2026-05-18T12:33:24Z"}],"graph_snapshots":[{"event_id":"sha256:5be1ca24fbf9555ab3e16fff9824d06ac6748674f2fc591fcd219260b6df6e87","target":"graph","created_at":"2026-05-17T23:43:42Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"While most previous work has focused on different pretraining objectives and architectures for transfer learning, we ask how to best adapt the pretrained model to a given target task. We focus on the two most common forms of adaptation, feature extraction (where the pretrained weights are frozen), and directly fine-tuning the pretrained model. Our empirical results across diverse NLP tasks with two state-of-the-art models show that the relative performance of fine-tuning vs. feature extraction depends on the similarity of the pretraining and target tasks. We explore possible explanations for t","authors_text":"Matthew E. Peters, Noah A. Smith, Sebastian Ruder","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-03-14T13:32:31Z","title":"To Tune or Not to Tune? Adapting Pretrained Representations to Diverse Tasks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1903.05987","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:0f875ad953a1720e6e63bdbbd9818343bb3b0a12aa702e4388bdecd0a45b975e","target":"record","created_at":"2026-05-17T23:43:42Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"4e209be446333f84609c163763e4312a1fe82b7d345bb2fbf2523d53fe7eb7cf","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-03-14T13:32:31Z","title_canon_sha256":"ebf111e46ced68405a524bb3a5babf72a466b66b88e91660d52563ba3f650f17"},"schema_version":"1.0","source":{"id":"1903.05987","kind":"arxiv","version":2}},"canonical_sha256":"6b644553b617a07eccf11738400eab49694329c8be0cf1cac0f7cde49532f444","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6b644553b617a07eccf11738400eab49694329c8be0cf1cac0f7cde49532f444","first_computed_at":"2026-05-17T23:43:42.437617Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:43:42.437617Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8fqx0ArUdgcw/Da9mWUGyq5nQzxggGKuHw+ceEcNNLAc5xgzalBcsbX4cPkh+z3v9lzaywQZ5VA1EO/8EMBqCg==","signature_status":"signed_v1","signed_at":"2026-05-17T23:43:42.438252Z","signed_message":"canonical_sha256_bytes"},"source_id":"1903.05987","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0f875ad953a1720e6e63bdbbd9818343bb3b0a12aa702e4388bdecd0a45b975e","sha256:5be1ca24fbf9555ab3e16fff9824d06ac6748674f2fc591fcd219260b6df6e87"],"state_sha256":"bf7c18d0e29b5de658c9b76f5b38871ac6869d2de630ef759c17a2a654b934f4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YOECAUTb66S+GvC1Da1e3f3/OQ9YVqQpWjAmwvUQuA+z7z7L3EnDX0xCQHh2kNNLOqrVVGvQF6ZBFw4CtKZxCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-29T14:37:00.021789Z","bundle_sha256":"bd0c742758de5590347caa63e08bf77d87fb4c94e5310cfc8e53c77d71a149c7"}}