{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RFKTTDNBIQGLYYRO3RDGDWUVVM","short_pith_number":"pith:RFKTTDNB","schema_version":"1.0","canonical_sha256":"8955398da1440cbc622edc4661da95ab30588c1691998f34d09165fbab004123","source":{"kind":"arxiv","id":"2506.18193","version":2},"attestation_state":"computed","paper":{"title":"DeInfoReg: A Decoupled Learning Framework for Better Training Throughput","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.DC"],"primary_cat":"cs.LG","authors_text":"Hung-Hsuan Chen, You-Teng Lin, Zih-Hao Huang","submitted_at":"2025-06-22T22:50:06Z","abstract_excerpt":"This paper introduces Decoupled Supervised Learning with Information Regularization (DeInfoReg), a novel approach that transforms a long gradient flow into multiple shorter ones, thereby mitigating the vanishing gradient problem. Integrating a pipeline strategy, DeInfoReg enables model parallelization across multiple GPUs, significantly improving training throughput. We compare our proposed method with standard backpropagation and other gradient flow decomposition techniques. Extensive experiments on diverse tasks and datasets demonstrate that DeInfoReg achieves superior performance and better"},"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":"2506.18193","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-22T22:50:06Z","cross_cats_sorted":["cs.AI","cs.DC"],"title_canon_sha256":"9312c54355b7f6c1fed68838f6b497f98dfb64c1705ffa47a4d31361e77896ec","abstract_canon_sha256":"8cb93895ef5306d4c453e7373c31cf2e665d17c42c8436571ea4899da575ef31"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:37:24.097294Z","signature_b64":"BNfOubH42wxE9vUsAkq+tAkXMfZcVhcgLmCbnP5us5pdRaNysZjMXD5H0NM/viGqhOSc38U7EsHKrw9TIICDCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8955398da1440cbc622edc4661da95ab30588c1691998f34d09165fbab004123","last_reissued_at":"2026-07-05T11:37:24.096667Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:37:24.096667Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DeInfoReg: A Decoupled Learning Framework for Better Training Throughput","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.DC"],"primary_cat":"cs.LG","authors_text":"Hung-Hsuan Chen, You-Teng Lin, Zih-Hao Huang","submitted_at":"2025-06-22T22:50:06Z","abstract_excerpt":"This paper introduces Decoupled Supervised Learning with Information Regularization (DeInfoReg), a novel approach that transforms a long gradient flow into multiple shorter ones, thereby mitigating the vanishing gradient problem. Integrating a pipeline strategy, DeInfoReg enables model parallelization across multiple GPUs, significantly improving training throughput. We compare our proposed method with standard backpropagation and other gradient flow decomposition techniques. Extensive experiments on diverse tasks and datasets demonstrate that DeInfoReg achieves superior performance and better"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.18193","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/2506.18193/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":"2506.18193","created_at":"2026-07-05T11:37:24.096742+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.18193v2","created_at":"2026-07-05T11:37:24.096742+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.18193","created_at":"2026-07-05T11:37:24.096742+00:00"},{"alias_kind":"pith_short_12","alias_value":"RFKTTDNBIQGL","created_at":"2026-07-05T11:37:24.096742+00:00"},{"alias_kind":"pith_short_16","alias_value":"RFKTTDNBIQGLYYRO","created_at":"2026-07-05T11:37:24.096742+00:00"},{"alias_kind":"pith_short_8","alias_value":"RFKTTDNB","created_at":"2026-07-05T11:37:24.096742+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/RFKTTDNBIQGLYYRO3RDGDWUVVM","json":"https://pith.science/pith/RFKTTDNBIQGLYYRO3RDGDWUVVM.json","graph_json":"https://pith.science/api/pith-number/RFKTTDNBIQGLYYRO3RDGDWUVVM/graph.json","events_json":"https://pith.science/api/pith-number/RFKTTDNBIQGLYYRO3RDGDWUVVM/events.json","paper":"https://pith.science/paper/RFKTTDNB"},"agent_actions":{"view_html":"https://pith.science/pith/RFKTTDNBIQGLYYRO3RDGDWUVVM","download_json":"https://pith.science/pith/RFKTTDNBIQGLYYRO3RDGDWUVVM.json","view_paper":"https://pith.science/paper/RFKTTDNB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.18193&json=true","fetch_graph":"https://pith.science/api/pith-number/RFKTTDNBIQGLYYRO3RDGDWUVVM/graph.json","fetch_events":"https://pith.science/api/pith-number/RFKTTDNBIQGLYYRO3RDGDWUVVM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RFKTTDNBIQGLYYRO3RDGDWUVVM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RFKTTDNBIQGLYYRO3RDGDWUVVM/action/storage_attestation","attest_author":"https://pith.science/pith/RFKTTDNBIQGLYYRO3RDGDWUVVM/action/author_attestation","sign_citation":"https://pith.science/pith/RFKTTDNBIQGLYYRO3RDGDWUVVM/action/citation_signature","submit_replication":"https://pith.science/pith/RFKTTDNBIQGLYYRO3RDGDWUVVM/action/replication_record"}},"created_at":"2026-07-05T11:37:24.096742+00:00","updated_at":"2026-07-05T11:37:24.096742+00:00"}