{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:GF67JE2NPTU33DUE7AQRKSOVFA","short_pith_number":"pith:GF67JE2N","schema_version":"1.0","canonical_sha256":"317df4934d7ce9bd8e84f8211549d5280ad7ff0f2a1c7707db41e62cd6694953","source":{"kind":"arxiv","id":"2211.13332","version":1},"attestation_state":"computed","paper":{"title":"Learning Compact Features via In-Training Representation Alignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Deng Pan, Dongxiao Zhu, Xiangrui Li, Xin Li, Yao Qiang","submitted_at":"2022-11-23T22:23:22Z","abstract_excerpt":"Deep neural networks (DNNs) for supervised learning can be viewed as a pipeline of the feature extractor (i.e., last hidden layer) and a linear classifier (i.e., output layer) that are trained jointly with stochastic gradient descent (SGD) on the loss function (e.g., cross-entropy). In each epoch, the true gradient of the loss function is estimated using a mini-batch sampled from the training set and model parameters are then updated with the mini-batch gradients. Although the latter provides an unbiased estimation of the former, they are subject to substantial variances derived from the size "},"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":"2211.13332","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-11-23T22:23:22Z","cross_cats_sorted":[],"title_canon_sha256":"8714ab057617b139ed2268858517ee9a8c96b4763f5b54e6323a2fcc2698ecd3","abstract_canon_sha256":"adb9fe6ca4a5c397de92e02371679a74255aa877ecb62b6faa975d7228a734d6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:19:02.521369Z","signature_b64":"gPSaN3SaqMWIOfyZmnb/HBcF0E5UDmUEyeEUyzSFmVdyYaKGhUmUBDJsi8gM7rL4PK8WE4+F0EQl5/n9TPAHBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"317df4934d7ce9bd8e84f8211549d5280ad7ff0f2a1c7707db41e62cd6694953","last_reissued_at":"2026-07-05T05:19:02.520885Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:19:02.520885Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Compact Features via In-Training Representation Alignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Deng Pan, Dongxiao Zhu, Xiangrui Li, Xin Li, Yao Qiang","submitted_at":"2022-11-23T22:23:22Z","abstract_excerpt":"Deep neural networks (DNNs) for supervised learning can be viewed as a pipeline of the feature extractor (i.e., last hidden layer) and a linear classifier (i.e., output layer) that are trained jointly with stochastic gradient descent (SGD) on the loss function (e.g., cross-entropy). In each epoch, the true gradient of the loss function is estimated using a mini-batch sampled from the training set and model parameters are then updated with the mini-batch gradients. Although the latter provides an unbiased estimation of the former, they are subject to substantial variances derived from the size "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.13332","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/2211.13332/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":"2211.13332","created_at":"2026-07-05T05:19:02.520960+00:00"},{"alias_kind":"arxiv_version","alias_value":"2211.13332v1","created_at":"2026-07-05T05:19:02.520960+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.13332","created_at":"2026-07-05T05:19:02.520960+00:00"},{"alias_kind":"pith_short_12","alias_value":"GF67JE2NPTU3","created_at":"2026-07-05T05:19:02.520960+00:00"},{"alias_kind":"pith_short_16","alias_value":"GF67JE2NPTU33DUE","created_at":"2026-07-05T05:19:02.520960+00:00"},{"alias_kind":"pith_short_8","alias_value":"GF67JE2N","created_at":"2026-07-05T05:19:02.520960+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/GF67JE2NPTU33DUE7AQRKSOVFA","json":"https://pith.science/pith/GF67JE2NPTU33DUE7AQRKSOVFA.json","graph_json":"https://pith.science/api/pith-number/GF67JE2NPTU33DUE7AQRKSOVFA/graph.json","events_json":"https://pith.science/api/pith-number/GF67JE2NPTU33DUE7AQRKSOVFA/events.json","paper":"https://pith.science/paper/GF67JE2N"},"agent_actions":{"view_html":"https://pith.science/pith/GF67JE2NPTU33DUE7AQRKSOVFA","download_json":"https://pith.science/pith/GF67JE2NPTU33DUE7AQRKSOVFA.json","view_paper":"https://pith.science/paper/GF67JE2N","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2211.13332&json=true","fetch_graph":"https://pith.science/api/pith-number/GF67JE2NPTU33DUE7AQRKSOVFA/graph.json","fetch_events":"https://pith.science/api/pith-number/GF67JE2NPTU33DUE7AQRKSOVFA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GF67JE2NPTU33DUE7AQRKSOVFA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GF67JE2NPTU33DUE7AQRKSOVFA/action/storage_attestation","attest_author":"https://pith.science/pith/GF67JE2NPTU33DUE7AQRKSOVFA/action/author_attestation","sign_citation":"https://pith.science/pith/GF67JE2NPTU33DUE7AQRKSOVFA/action/citation_signature","submit_replication":"https://pith.science/pith/GF67JE2NPTU33DUE7AQRKSOVFA/action/replication_record"}},"created_at":"2026-07-05T05:19:02.520960+00:00","updated_at":"2026-07-05T05:19:02.520960+00:00"}