{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:JTSSYMTWECNS7URGS3Y4ZCQ6FK","short_pith_number":"pith:JTSSYMTW","schema_version":"1.0","canonical_sha256":"4ce52c3276209b2fd22696f1cc8a1e2aa7cf77488a4fb5e11a830722c8862c4a","source":{"kind":"arxiv","id":"2504.02211","version":2},"attestation_state":"computed","paper":{"title":"FT-Transformer: Resilient and Reliable Transformer with End-to-End Fault Tolerant Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.DC","authors_text":"Haiyang Hu, Huangliang Dai, Jiajun Huang, Shixun Wu, Yue Zhu, Zizhe Jian, Zizhong Chen","submitted_at":"2025-04-03T02:05:08Z","abstract_excerpt":"Transformer models rely on High-Performance Computing (HPC) resources for inference, where soft errors are inevitable in large-scale systems, making the reliability of the model particularly critical. Existing fault tolerance frameworks for Transformers are designed at the operation level without architectural optimization, leading to significant computational and memory overhead, which in turn reduces protection efficiency and limits scalability to larger models. In this paper, we implement module-level protection for Transformers by treating the operations within the attention module as a si"},"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":"2504.02211","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2025-04-03T02:05:08Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"ae31f127712b86765c687c5d14e2813c9d5d5028150a352993b3b19acf30103f","abstract_canon_sha256":"297bb1270ae8747588cec31daf83f00b3d5812fcd3f974d999d63f802aa984a1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:53:00.910922Z","signature_b64":"orMJhIY8O2igL3Fuz63OWbH+1RvDbfKmnDLy1CcKILtzFrE++mHz03huWpwq/pqWax0aXnaQdXW4+7E6JNErAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4ce52c3276209b2fd22696f1cc8a1e2aa7cf77488a4fb5e11a830722c8862c4a","last_reissued_at":"2026-07-05T11:53:00.910458Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:53:00.910458Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"FT-Transformer: Resilient and Reliable Transformer with End-to-End Fault Tolerant Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.DC","authors_text":"Haiyang Hu, Huangliang Dai, Jiajun Huang, Shixun Wu, Yue Zhu, Zizhe Jian, Zizhong Chen","submitted_at":"2025-04-03T02:05:08Z","abstract_excerpt":"Transformer models rely on High-Performance Computing (HPC) resources for inference, where soft errors are inevitable in large-scale systems, making the reliability of the model particularly critical. Existing fault tolerance frameworks for Transformers are designed at the operation level without architectural optimization, leading to significant computational and memory overhead, which in turn reduces protection efficiency and limits scalability to larger models. In this paper, we implement module-level protection for Transformers by treating the operations within the attention module as a si"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.02211","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/2504.02211/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":"2504.02211","created_at":"2026-07-05T11:53:00.910515+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.02211v2","created_at":"2026-07-05T11:53:00.910515+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.02211","created_at":"2026-07-05T11:53:00.910515+00:00"},{"alias_kind":"pith_short_12","alias_value":"JTSSYMTWECNS","created_at":"2026-07-05T11:53:00.910515+00:00"},{"alias_kind":"pith_short_16","alias_value":"JTSSYMTWECNS7URG","created_at":"2026-07-05T11:53:00.910515+00:00"},{"alias_kind":"pith_short_8","alias_value":"JTSSYMTW","created_at":"2026-07-05T11:53:00.910515+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.27934","citing_title":"Self-Verifying Measurement Records: Hash-Linked Evidence Graphs for Hardware Benchmarking","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.28118","citing_title":"Hierarchical Fault Detection and Diagnosis for Transformer Architectures","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.28118","citing_title":"Hierarchical Fault Detection and Diagnosis for Transformer Architectures","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08973","citing_title":"Tight Lower Bounds on The Single-Error Detection Threshold for Analog Error-Correcting Codes","ref_index":26,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JTSSYMTWECNS7URGS3Y4ZCQ6FK","json":"https://pith.science/pith/JTSSYMTWECNS7URGS3Y4ZCQ6FK.json","graph_json":"https://pith.science/api/pith-number/JTSSYMTWECNS7URGS3Y4ZCQ6FK/graph.json","events_json":"https://pith.science/api/pith-number/JTSSYMTWECNS7URGS3Y4ZCQ6FK/events.json","paper":"https://pith.science/paper/JTSSYMTW"},"agent_actions":{"view_html":"https://pith.science/pith/JTSSYMTWECNS7URGS3Y4ZCQ6FK","download_json":"https://pith.science/pith/JTSSYMTWECNS7URGS3Y4ZCQ6FK.json","view_paper":"https://pith.science/paper/JTSSYMTW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.02211&json=true","fetch_graph":"https://pith.science/api/pith-number/JTSSYMTWECNS7URGS3Y4ZCQ6FK/graph.json","fetch_events":"https://pith.science/api/pith-number/JTSSYMTWECNS7URGS3Y4ZCQ6FK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JTSSYMTWECNS7URGS3Y4ZCQ6FK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JTSSYMTWECNS7URGS3Y4ZCQ6FK/action/storage_attestation","attest_author":"https://pith.science/pith/JTSSYMTWECNS7URGS3Y4ZCQ6FK/action/author_attestation","sign_citation":"https://pith.science/pith/JTSSYMTWECNS7URGS3Y4ZCQ6FK/action/citation_signature","submit_replication":"https://pith.science/pith/JTSSYMTWECNS7URGS3Y4ZCQ6FK/action/replication_record"}},"created_at":"2026-07-05T11:53:00.910515+00:00","updated_at":"2026-07-05T11:53:00.910515+00:00"}