{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:QPLKOJ52PVJNS3QMNDDOQ3CM4C","short_pith_number":"pith:QPLKOJ52","canonical_record":{"source":{"id":"2105.03791","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-05-08T22:31:51Z","cross_cats_sorted":[],"title_canon_sha256":"9caec14a94bceb2fcba5b08dffd125ea5e2b9cac8aed5e94f551037ff9feaab5","abstract_canon_sha256":"8a8c5ed31470129a570a1f8560a3c815cd1f67b0c4b2113ddd7f7a19b3a6bf63"},"schema_version":"1.0"},"canonical_sha256":"83d6a727ba7d52d96e0c68c6e86c4ce0870c3502e2b47625acf6c047cf6610ba","source":{"kind":"arxiv","id":"2105.03791","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2105.03791","created_at":"2026-07-05T04:56:10Z"},{"alias_kind":"arxiv_version","alias_value":"2105.03791v2","created_at":"2026-07-05T04:56:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2105.03791","created_at":"2026-07-05T04:56:10Z"},{"alias_kind":"pith_short_12","alias_value":"QPLKOJ52PVJN","created_at":"2026-07-05T04:56:10Z"},{"alias_kind":"pith_short_16","alias_value":"QPLKOJ52PVJNS3QM","created_at":"2026-07-05T04:56:10Z"},{"alias_kind":"pith_short_8","alias_value":"QPLKOJ52","created_at":"2026-07-05T04:56:10Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:QPLKOJ52PVJNS3QMNDDOQ3CM4C","target":"record","payload":{"canonical_record":{"source":{"id":"2105.03791","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-05-08T22:31:51Z","cross_cats_sorted":[],"title_canon_sha256":"9caec14a94bceb2fcba5b08dffd125ea5e2b9cac8aed5e94f551037ff9feaab5","abstract_canon_sha256":"8a8c5ed31470129a570a1f8560a3c815cd1f67b0c4b2113ddd7f7a19b3a6bf63"},"schema_version":"1.0"},"canonical_sha256":"83d6a727ba7d52d96e0c68c6e86c4ce0870c3502e2b47625acf6c047cf6610ba","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:56:10.773643Z","signature_b64":"rBYKrcxwIyc5ygBak3PKpoi1ZQ53Oe92b8SraQN3ZZeL+EehuIFzEfW1wxIKzX5dNXve6hqPtw0deFjNrpbIDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"83d6a727ba7d52d96e0c68c6e86c4ce0870c3502e2b47625acf6c047cf6610ba","last_reissued_at":"2026-07-05T04:56:10.773251Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:56:10.773251Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2105.03791","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":"2026-07-05T04:56:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kQkmCEPVYtzyE846cP62wq9oRVg8Rg2BR/Xs1JLwUjJERgOJMrYGU5H+bUoWjcOjlQ1T4dPrOz+IO5gG6yYVAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-01T12:04:46.510948Z"},"content_sha256":"e9dda9851781f24ce85982d203bbb86e0aa369a21f41963b9533066e9fb44762","schema_version":"1.0","event_id":"sha256:e9dda9851781f24ce85982d203bbb86e0aa369a21f41963b9533066e9fb44762"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:QPLKOJ52PVJNS3QMNDDOQ3CM4C","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Enhancing Transformers with Gradient Boosted Decision Trees for NLI Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Benjamin Minixhofer, Ignacio Iacobacci, Milan Gritta","submitted_at":"2021-05-08T22:31:51Z","abstract_excerpt":"Transfer learning has become the dominant paradigm for many natural language processing tasks. In addition to models being pretrained on large datasets, they can be further trained on intermediate (supervised) tasks that are similar to the target task. For small Natural Language Inference (NLI) datasets, language modelling is typically followed by pretraining on a large (labelled) NLI dataset before fine-tuning with each NLI subtask. In this work, we explore Gradient Boosted Decision Trees (GBDTs) as an alternative to the commonly used Multi-Layer Perceptron (MLP) classification head. GBDTs ha"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2105.03791","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/2105.03791/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T04:56:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"09TVBhjUl7WYsK1d11pgz+2YB8leONvFjxiPVVgeC0+GwjtrCAbfzKYx72EOEXeLa+l/N5W3JW3RsgGYb7TtAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-01T12:04:46.511449Z"},"content_sha256":"57b0bdbe66b0ab7a42c676fcdb7f65347bbcf98fb5bf1d9a93bec932d177b176","schema_version":"1.0","event_id":"sha256:57b0bdbe66b0ab7a42c676fcdb7f65347bbcf98fb5bf1d9a93bec932d177b176"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/QPLKOJ52PVJNS3QMNDDOQ3CM4C/bundle.json","state_url":"https://pith.science/pith/QPLKOJ52PVJNS3QMNDDOQ3CM4C/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/QPLKOJ52PVJNS3QMNDDOQ3CM4C/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":"2026-08-01T12:04:46Z","links":{"resolver":"https://pith.science/pith/QPLKOJ52PVJNS3QMNDDOQ3CM4C","bundle":"https://pith.science/pith/QPLKOJ52PVJNS3QMNDDOQ3CM4C/bundle.json","state":"https://pith.science/pith/QPLKOJ52PVJNS3QMNDDOQ3CM4C/state.json","well_known_bundle":"https://pith.science/.well-known/pith/QPLKOJ52PVJNS3QMNDDOQ3CM4C/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:QPLKOJ52PVJNS3QMNDDOQ3CM4C","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":"8a8c5ed31470129a570a1f8560a3c815cd1f67b0c4b2113ddd7f7a19b3a6bf63","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-05-08T22:31:51Z","title_canon_sha256":"9caec14a94bceb2fcba5b08dffd125ea5e2b9cac8aed5e94f551037ff9feaab5"},"schema_version":"1.0","source":{"id":"2105.03791","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2105.03791","created_at":"2026-07-05T04:56:10Z"},{"alias_kind":"arxiv_version","alias_value":"2105.03791v2","created_at":"2026-07-05T04:56:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2105.03791","created_at":"2026-07-05T04:56:10Z"},{"alias_kind":"pith_short_12","alias_value":"QPLKOJ52PVJN","created_at":"2026-07-05T04:56:10Z"},{"alias_kind":"pith_short_16","alias_value":"QPLKOJ52PVJNS3QM","created_at":"2026-07-05T04:56:10Z"},{"alias_kind":"pith_short_8","alias_value":"QPLKOJ52","created_at":"2026-07-05T04:56:10Z"}],"graph_snapshots":[{"event_id":"sha256:57b0bdbe66b0ab7a42c676fcdb7f65347bbcf98fb5bf1d9a93bec932d177b176","target":"graph","created_at":"2026-07-05T04:56:10Z","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"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2105.03791/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Transfer learning has become the dominant paradigm for many natural language processing tasks. In addition to models being pretrained on large datasets, they can be further trained on intermediate (supervised) tasks that are similar to the target task. For small Natural Language Inference (NLI) datasets, language modelling is typically followed by pretraining on a large (labelled) NLI dataset before fine-tuning with each NLI subtask. In this work, we explore Gradient Boosted Decision Trees (GBDTs) as an alternative to the commonly used Multi-Layer Perceptron (MLP) classification head. GBDTs ha","authors_text":"Benjamin Minixhofer, Ignacio Iacobacci, Milan Gritta","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-05-08T22:31:51Z","title":"Enhancing Transformers with Gradient Boosted Decision Trees for NLI Fine-Tuning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2105.03791","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:e9dda9851781f24ce85982d203bbb86e0aa369a21f41963b9533066e9fb44762","target":"record","created_at":"2026-07-05T04:56:10Z","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":"8a8c5ed31470129a570a1f8560a3c815cd1f67b0c4b2113ddd7f7a19b3a6bf63","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-05-08T22:31:51Z","title_canon_sha256":"9caec14a94bceb2fcba5b08dffd125ea5e2b9cac8aed5e94f551037ff9feaab5"},"schema_version":"1.0","source":{"id":"2105.03791","kind":"arxiv","version":2}},"canonical_sha256":"83d6a727ba7d52d96e0c68c6e86c4ce0870c3502e2b47625acf6c047cf6610ba","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"83d6a727ba7d52d96e0c68c6e86c4ce0870c3502e2b47625acf6c047cf6610ba","first_computed_at":"2026-07-05T04:56:10.773251Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T04:56:10.773251Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"rBYKrcxwIyc5ygBak3PKpoi1ZQ53Oe92b8SraQN3ZZeL+EehuIFzEfW1wxIKzX5dNXve6hqPtw0deFjNrpbIDw==","signature_status":"signed_v1","signed_at":"2026-07-05T04:56:10.773643Z","signed_message":"canonical_sha256_bytes"},"source_id":"2105.03791","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e9dda9851781f24ce85982d203bbb86e0aa369a21f41963b9533066e9fb44762","sha256:57b0bdbe66b0ab7a42c676fcdb7f65347bbcf98fb5bf1d9a93bec932d177b176"],"state_sha256":"e08c506f4b9dfe186c953023d7f55ac131e771ed6f77261b687ae43d8e0625c0"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"st1l4Ub6pA1MTgETTO4xCBhWzZFKF1AfXyOrFKkh1Jl5i1hlE6GR3FNimuZciT787El0S7M94DQ/1NGFrosOBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-01T12:04:46.517206Z","bundle_sha256":"0bbd792d35145f3eb49e3d3d69e225c43fb6e34b4f3f365bb1a39615c68b21d1"}}