{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:C3CAU45BP2T7ZHH7W2IGKM3EFA","short_pith_number":"pith:C3CAU45B","schema_version":"1.0","canonical_sha256":"16c40a73a17ea7fc9cffb6906533642801c0e9510ac93391262e33e9a308be5d","source":{"kind":"arxiv","id":"2205.01228","version":2},"attestation_state":"computed","paper":{"title":"Paragraph-based Transformer Pre-training for Multi-Sentence Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alessandro Moschitti, Luca Di Liello, Luca Soldaini, Siddhant Garg","submitted_at":"2022-05-02T21:41:14Z","abstract_excerpt":"Inference tasks such as answer sentence selection (AS2) or fact verification are typically solved by fine-tuning transformer-based models as individual sentence-pair classifiers. Recent studies show that these tasks benefit from modeling dependencies across multiple candidate sentences jointly. In this paper, we first show that popular pre-trained transformers perform poorly when used for fine-tuning on multi-candidate inference tasks. We then propose a new pre-training objective that models the paragraph-level semantics across multiple input sentences. Our evaluation on three AS2 and one fact"},"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":"2205.01228","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-05-02T21:41:14Z","cross_cats_sorted":[],"title_canon_sha256":"2d9a43d0c42f1dd0a1285c11e25f189c475a5900708e4a2d5ff4b7c8757f521b","abstract_canon_sha256":"ef975f33203730ad8c74efcddea89e284d6a507b363cbf3c542e669b2d4420ad"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:38:14.776209Z","signature_b64":"YHeuGEIiusN146IPIFteWrfOb7kBUwnzQJ0m4PuET/tYos9G8OV6T5ut8Z88MI5dk8uOHV3kHWNnjX7aWu2+CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"16c40a73a17ea7fc9cffb6906533642801c0e9510ac93391262e33e9a308be5d","last_reissued_at":"2026-07-05T04:38:14.775777Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:38:14.775777Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Paragraph-based Transformer Pre-training for Multi-Sentence Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alessandro Moschitti, Luca Di Liello, Luca Soldaini, Siddhant Garg","submitted_at":"2022-05-02T21:41:14Z","abstract_excerpt":"Inference tasks such as answer sentence selection (AS2) or fact verification are typically solved by fine-tuning transformer-based models as individual sentence-pair classifiers. Recent studies show that these tasks benefit from modeling dependencies across multiple candidate sentences jointly. In this paper, we first show that popular pre-trained transformers perform poorly when used for fine-tuning on multi-candidate inference tasks. We then propose a new pre-training objective that models the paragraph-level semantics across multiple input sentences. Our evaluation on three AS2 and one fact"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.01228","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/2205.01228/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":"2205.01228","created_at":"2026-07-05T04:38:14.775838+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.01228v2","created_at":"2026-07-05T04:38:14.775838+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.01228","created_at":"2026-07-05T04:38:14.775838+00:00"},{"alias_kind":"pith_short_12","alias_value":"C3CAU45BP2T7","created_at":"2026-07-05T04:38:14.775838+00:00"},{"alias_kind":"pith_short_16","alias_value":"C3CAU45BP2T7ZHH7","created_at":"2026-07-05T04:38:14.775838+00:00"},{"alias_kind":"pith_short_8","alias_value":"C3CAU45B","created_at":"2026-07-05T04:38:14.775838+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/C3CAU45BP2T7ZHH7W2IGKM3EFA","json":"https://pith.science/pith/C3CAU45BP2T7ZHH7W2IGKM3EFA.json","graph_json":"https://pith.science/api/pith-number/C3CAU45BP2T7ZHH7W2IGKM3EFA/graph.json","events_json":"https://pith.science/api/pith-number/C3CAU45BP2T7ZHH7W2IGKM3EFA/events.json","paper":"https://pith.science/paper/C3CAU45B"},"agent_actions":{"view_html":"https://pith.science/pith/C3CAU45BP2T7ZHH7W2IGKM3EFA","download_json":"https://pith.science/pith/C3CAU45BP2T7ZHH7W2IGKM3EFA.json","view_paper":"https://pith.science/paper/C3CAU45B","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.01228&json=true","fetch_graph":"https://pith.science/api/pith-number/C3CAU45BP2T7ZHH7W2IGKM3EFA/graph.json","fetch_events":"https://pith.science/api/pith-number/C3CAU45BP2T7ZHH7W2IGKM3EFA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/C3CAU45BP2T7ZHH7W2IGKM3EFA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/C3CAU45BP2T7ZHH7W2IGKM3EFA/action/storage_attestation","attest_author":"https://pith.science/pith/C3CAU45BP2T7ZHH7W2IGKM3EFA/action/author_attestation","sign_citation":"https://pith.science/pith/C3CAU45BP2T7ZHH7W2IGKM3EFA/action/citation_signature","submit_replication":"https://pith.science/pith/C3CAU45BP2T7ZHH7W2IGKM3EFA/action/replication_record"}},"created_at":"2026-07-05T04:38:14.775838+00:00","updated_at":"2026-07-05T04:38:14.775838+00:00"}