{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:5YAJ3RYIFARXPFFAB2CUGHGSLU","short_pith_number":"pith:5YAJ3RYI","schema_version":"1.0","canonical_sha256":"ee009dc70828237794a00e85431cd25d3db14f5c819dfaa07c8304253dc47462","source":{"kind":"arxiv","id":"2305.14561","version":5},"attestation_state":"computed","paper":{"title":"NeFT: Negative Feedback Training to Improve Robustness of Compute-In-Memory DNN Accelerators","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.AR"],"primary_cat":"cs.LG","authors_text":"Dailin Gan, Jun Xia, Wujie Wen, Xiaobo Sharon Hu, Yifan Qin, Yiyu Shi, Zheyu Yan, Zixuan Pan","submitted_at":"2023-05-23T22:56:26Z","abstract_excerpt":"Compute-in-memory accelerators built upon non-volatile memory devices excel in energy efficiency and latency when performing deep neural network (DNN) inference, thanks to their in-situ data processing capability. However, the stochastic nature and intrinsic variations of non-volatile memory devices often result in performance degradation during DNN inference. Introducing these non-ideal device behaviors in DNN training enhances robustness, but drawbacks include limited accuracy improvement, reduced prediction confidence, and convergence issues. This arises from a mismatch between the determin"},"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":"2305.14561","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-05-23T22:56:26Z","cross_cats_sorted":["cs.AI","cs.AR"],"title_canon_sha256":"339ee012154d84498843cba8645e1cc8d46e80d6c5d86665e42f8e99caeacf82","abstract_canon_sha256":"91e176c9ba02c4d2db50569f18929174a8f6a82f8c60cc44bc14da5162e27124"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:54:37.837444Z","signature_b64":"qypBdeJrKgXlwmW3tqe91MkddEvZBnwGI4V0AeeGRkWpex8C+QbWIznDeqlqz7sp8vpv8sCFUS6hO2W/fbXPAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ee009dc70828237794a00e85431cd25d3db14f5c819dfaa07c8304253dc47462","last_reissued_at":"2026-07-05T11:54:37.836938Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:54:37.836938Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"NeFT: Negative Feedback Training to Improve Robustness of Compute-In-Memory DNN Accelerators","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.AR"],"primary_cat":"cs.LG","authors_text":"Dailin Gan, Jun Xia, Wujie Wen, Xiaobo Sharon Hu, Yifan Qin, Yiyu Shi, Zheyu Yan, Zixuan Pan","submitted_at":"2023-05-23T22:56:26Z","abstract_excerpt":"Compute-in-memory accelerators built upon non-volatile memory devices excel in energy efficiency and latency when performing deep neural network (DNN) inference, thanks to their in-situ data processing capability. However, the stochastic nature and intrinsic variations of non-volatile memory devices often result in performance degradation during DNN inference. Introducing these non-ideal device behaviors in DNN training enhances robustness, but drawbacks include limited accuracy improvement, reduced prediction confidence, and convergence issues. This arises from a mismatch between the determin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.14561","kind":"arxiv","version":5},"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/2305.14561/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":"2305.14561","created_at":"2026-07-05T11:54:37.836997+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.14561v5","created_at":"2026-07-05T11:54:37.836997+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.14561","created_at":"2026-07-05T11:54:37.836997+00:00"},{"alias_kind":"pith_short_12","alias_value":"5YAJ3RYIFARX","created_at":"2026-07-05T11:54:37.836997+00:00"},{"alias_kind":"pith_short_16","alias_value":"5YAJ3RYIFARXPFFA","created_at":"2026-07-05T11:54:37.836997+00:00"},{"alias_kind":"pith_short_8","alias_value":"5YAJ3RYI","created_at":"2026-07-05T11:54:37.836997+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.09588","citing_title":"Atleus: Accelerating Transformers on the Edge Enabled by 3D Heterogeneous Manycore Architectures","ref_index":54,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5YAJ3RYIFARXPFFAB2CUGHGSLU","json":"https://pith.science/pith/5YAJ3RYIFARXPFFAB2CUGHGSLU.json","graph_json":"https://pith.science/api/pith-number/5YAJ3RYIFARXPFFAB2CUGHGSLU/graph.json","events_json":"https://pith.science/api/pith-number/5YAJ3RYIFARXPFFAB2CUGHGSLU/events.json","paper":"https://pith.science/paper/5YAJ3RYI"},"agent_actions":{"view_html":"https://pith.science/pith/5YAJ3RYIFARXPFFAB2CUGHGSLU","download_json":"https://pith.science/pith/5YAJ3RYIFARXPFFAB2CUGHGSLU.json","view_paper":"https://pith.science/paper/5YAJ3RYI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.14561&json=true","fetch_graph":"https://pith.science/api/pith-number/5YAJ3RYIFARXPFFAB2CUGHGSLU/graph.json","fetch_events":"https://pith.science/api/pith-number/5YAJ3RYIFARXPFFAB2CUGHGSLU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5YAJ3RYIFARXPFFAB2CUGHGSLU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5YAJ3RYIFARXPFFAB2CUGHGSLU/action/storage_attestation","attest_author":"https://pith.science/pith/5YAJ3RYIFARXPFFAB2CUGHGSLU/action/author_attestation","sign_citation":"https://pith.science/pith/5YAJ3RYIFARXPFFAB2CUGHGSLU/action/citation_signature","submit_replication":"https://pith.science/pith/5YAJ3RYIFARXPFFAB2CUGHGSLU/action/replication_record"}},"created_at":"2026-07-05T11:54:37.836997+00:00","updated_at":"2026-07-05T11:54:37.836997+00:00"}