{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:35SUQBVLFRVKJBJ6PDD4VCVOYJ","short_pith_number":"pith:35SUQBVL","schema_version":"1.0","canonical_sha256":"df654806ab2c6aa4853e78c7ca8aaec25d36704abd57e12c83b47e84c0f3a417","source":{"kind":"arxiv","id":"2404.06911","version":1},"attestation_state":"computed","paper":{"title":"GraSAME: Injecting Token-Level Structural Information to Pretrained Language Models via Graph-guided Self-Attention Mechanism","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Michael F\\\"arber, Shuzhou Yuan","submitted_at":"2024-04-10T11:03:57Z","abstract_excerpt":"Pretrained Language Models (PLMs) benefit from external knowledge stored in graph structures for various downstream tasks. However, bridging the modality gap between graph structures and text remains a significant challenge. Traditional methods like linearizing graphs for PLMs lose vital graph connectivity, whereas Graph Neural Networks (GNNs) require cumbersome processes for integration into PLMs. In this work, we propose a novel graph-guided self-attention mechanism, GraSAME. GraSAME seamlessly incorporates token-level structural information into PLMs without necessitating additional alignme"},"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":"2404.06911","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-10T11:03:57Z","cross_cats_sorted":[],"title_canon_sha256":"8146c7f801923333b854e133a3687b60c06df4c995af6e5b6ee2e2ddf42bdacd","abstract_canon_sha256":"6d9bac6cc6219fd8596a865955b97499e2552cd3ce1f3acf5f61695d7749ebcb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:06:32.098799Z","signature_b64":"+HGk5VRPWbsJE8kMpe848iVSmhJtjTXvHTS6VK/uQuJ0bV+YCFGEzxBaFsU8KIDtRTGmPs//xITgBL+2ezbdBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"df654806ab2c6aa4853e78c7ca8aaec25d36704abd57e12c83b47e84c0f3a417","last_reissued_at":"2026-07-05T08:06:32.098394Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:06:32.098394Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GraSAME: Injecting Token-Level Structural Information to Pretrained Language Models via Graph-guided Self-Attention Mechanism","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Michael F\\\"arber, Shuzhou Yuan","submitted_at":"2024-04-10T11:03:57Z","abstract_excerpt":"Pretrained Language Models (PLMs) benefit from external knowledge stored in graph structures for various downstream tasks. However, bridging the modality gap between graph structures and text remains a significant challenge. Traditional methods like linearizing graphs for PLMs lose vital graph connectivity, whereas Graph Neural Networks (GNNs) require cumbersome processes for integration into PLMs. In this work, we propose a novel graph-guided self-attention mechanism, GraSAME. GraSAME seamlessly incorporates token-level structural information into PLMs without necessitating additional alignme"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.06911","kind":"arxiv","version":1},"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/2404.06911/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":"2404.06911","created_at":"2026-07-05T08:06:32.098451+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.06911v1","created_at":"2026-07-05T08:06:32.098451+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.06911","created_at":"2026-07-05T08:06:32.098451+00:00"},{"alias_kind":"pith_short_12","alias_value":"35SUQBVLFRVK","created_at":"2026-07-05T08:06:32.098451+00:00"},{"alias_kind":"pith_short_16","alias_value":"35SUQBVLFRVKJBJ6","created_at":"2026-07-05T08:06:32.098451+00:00"},{"alias_kind":"pith_short_8","alias_value":"35SUQBVL","created_at":"2026-07-05T08:06:32.098451+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/35SUQBVLFRVKJBJ6PDD4VCVOYJ","json":"https://pith.science/pith/35SUQBVLFRVKJBJ6PDD4VCVOYJ.json","graph_json":"https://pith.science/api/pith-number/35SUQBVLFRVKJBJ6PDD4VCVOYJ/graph.json","events_json":"https://pith.science/api/pith-number/35SUQBVLFRVKJBJ6PDD4VCVOYJ/events.json","paper":"https://pith.science/paper/35SUQBVL"},"agent_actions":{"view_html":"https://pith.science/pith/35SUQBVLFRVKJBJ6PDD4VCVOYJ","download_json":"https://pith.science/pith/35SUQBVLFRVKJBJ6PDD4VCVOYJ.json","view_paper":"https://pith.science/paper/35SUQBVL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.06911&json=true","fetch_graph":"https://pith.science/api/pith-number/35SUQBVLFRVKJBJ6PDD4VCVOYJ/graph.json","fetch_events":"https://pith.science/api/pith-number/35SUQBVLFRVKJBJ6PDD4VCVOYJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/35SUQBVLFRVKJBJ6PDD4VCVOYJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/35SUQBVLFRVKJBJ6PDD4VCVOYJ/action/storage_attestation","attest_author":"https://pith.science/pith/35SUQBVLFRVKJBJ6PDD4VCVOYJ/action/author_attestation","sign_citation":"https://pith.science/pith/35SUQBVLFRVKJBJ6PDD4VCVOYJ/action/citation_signature","submit_replication":"https://pith.science/pith/35SUQBVLFRVKJBJ6PDD4VCVOYJ/action/replication_record"}},"created_at":"2026-07-05T08:06:32.098451+00:00","updated_at":"2026-07-05T08:06:32.098451+00:00"}