{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:LYDYETUMUWEFXUXZR7BWRNDHJW","short_pith_number":"pith:LYDYETUM","schema_version":"1.0","canonical_sha256":"5e07824e8ca5885bd2f98fc368b4674d97f60bcf90c00af3f574ac189a64569d","source":{"kind":"arxiv","id":"2407.21515","version":2},"attestation_state":"computed","paper":{"title":"Learning Effective Representations for Retrieval Using Self-Distillation with Adaptive Relevance Margins","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Harrisen Scells, Lukas Gienapp, Martin Potthast, Niklas Deckers","submitted_at":"2024-07-31T10:33:32Z","abstract_excerpt":"Representation-based retrieval models, so-called bi-encoders, estimate the relevance of a document to a query by calculating the similarity of their respective embeddings. Current state-of-the-art bi-encoders are trained using an expensive training regime involving knowledge distillation from a teacher model and batch-sampling. Instead of relying on a teacher model, we contribute a novel parameter-free loss function for self-supervision that exploits the pre-trained language modeling capabilities of the encoder model as a training signal, eliminating the need for batch sampling by performing i"},"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":"2407.21515","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2024-07-31T10:33:32Z","cross_cats_sorted":[],"title_canon_sha256":"245d68cb7431a85b482857971360964b1bad2c9a9b47304427305e446c351844","abstract_canon_sha256":"cb960fc8a6be74560b7b5a894df88ad583c109c700b49ed4824a83d3d2e0f283"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:25:03.390278Z","signature_b64":"+FpSHOWiwb+zcCMcJsK24nvcx+vMNoT0MMdZiitPlds8BJd+GY2LidEvPedPQ2FWSdpYGFJEycBObfBN+35NDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5e07824e8ca5885bd2f98fc368b4674d97f60bcf90c00af3f574ac189a64569d","last_reissued_at":"2026-07-05T11:25:03.388856Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:25:03.388856Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Effective Representations for Retrieval Using Self-Distillation with Adaptive Relevance Margins","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Harrisen Scells, Lukas Gienapp, Martin Potthast, Niklas Deckers","submitted_at":"2024-07-31T10:33:32Z","abstract_excerpt":"Representation-based retrieval models, so-called bi-encoders, estimate the relevance of a document to a query by calculating the similarity of their respective embeddings. Current state-of-the-art bi-encoders are trained using an expensive training regime involving knowledge distillation from a teacher model and batch-sampling. Instead of relying on a teacher model, we contribute a novel parameter-free loss function for self-supervision that exploits the pre-trained language modeling capabilities of the encoder model as a training signal, eliminating the need for batch sampling by performing i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.21515","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/2407.21515/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":"2407.21515","created_at":"2026-07-05T11:25:03.389764+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.21515v2","created_at":"2026-07-05T11:25:03.389764+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.21515","created_at":"2026-07-05T11:25:03.389764+00:00"},{"alias_kind":"pith_short_12","alias_value":"LYDYETUMUWEF","created_at":"2026-07-05T11:25:03.389764+00:00"},{"alias_kind":"pith_short_16","alias_value":"LYDYETUMUWEFXUXZ","created_at":"2026-07-05T11:25:03.389764+00:00"},{"alias_kind":"pith_short_8","alias_value":"LYDYETUM","created_at":"2026-07-05T11:25:03.389764+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/LYDYETUMUWEFXUXZR7BWRNDHJW","json":"https://pith.science/pith/LYDYETUMUWEFXUXZR7BWRNDHJW.json","graph_json":"https://pith.science/api/pith-number/LYDYETUMUWEFXUXZR7BWRNDHJW/graph.json","events_json":"https://pith.science/api/pith-number/LYDYETUMUWEFXUXZR7BWRNDHJW/events.json","paper":"https://pith.science/paper/LYDYETUM"},"agent_actions":{"view_html":"https://pith.science/pith/LYDYETUMUWEFXUXZR7BWRNDHJW","download_json":"https://pith.science/pith/LYDYETUMUWEFXUXZR7BWRNDHJW.json","view_paper":"https://pith.science/paper/LYDYETUM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.21515&json=true","fetch_graph":"https://pith.science/api/pith-number/LYDYETUMUWEFXUXZR7BWRNDHJW/graph.json","fetch_events":"https://pith.science/api/pith-number/LYDYETUMUWEFXUXZR7BWRNDHJW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LYDYETUMUWEFXUXZR7BWRNDHJW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LYDYETUMUWEFXUXZR7BWRNDHJW/action/storage_attestation","attest_author":"https://pith.science/pith/LYDYETUMUWEFXUXZR7BWRNDHJW/action/author_attestation","sign_citation":"https://pith.science/pith/LYDYETUMUWEFXUXZR7BWRNDHJW/action/citation_signature","submit_replication":"https://pith.science/pith/LYDYETUMUWEFXUXZR7BWRNDHJW/action/replication_record"}},"created_at":"2026-07-05T11:25:03.389764+00:00","updated_at":"2026-07-05T11:25:03.389764+00:00"}