{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:74MMAXNA7V5HMZ5NFWA6DBQI7X","short_pith_number":"pith:74MMAXNA","schema_version":"1.0","canonical_sha256":"ff18c05da0fd7a7667ad2d81e18608fdd245eb2e2a118dee597d214e7e3b12a5","source":{"kind":"arxiv","id":"2507.17953","version":1},"attestation_state":"computed","paper":{"title":"Clo-HDnn: A 4.66 TFLOPS/W and 3.78 TOPS/W Continual On-Device Learning Accelerator with Energy-efficient Hyperdimensional Computing via Progressive Search","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AR","authors_text":"Carlos H. Diaz, Chang Eun Song, Gert Cauwenberghs, Gopabandhu Hota, Haichao Yang, Keming Fan, Kerem Akarvardar, Leo Liu, Meng-Fan Chang, Mingu Kang, Soumil Jain, Tajana Rosing, Weihong Xu","submitted_at":"2025-07-23T21:50:28Z","abstract_excerpt":"Clo-HDnn is an on-device learning (ODL) accelerator designed for emerging continual learning (CL) tasks. Clo-HDnn integrates hyperdimensional computing (HDC) along with low-cost Kronecker HD Encoder and weight clustering feature extraction (WCFE) to optimize accuracy and efficiency. Clo-HDnn adopts gradient-free CL to efficiently update and store the learned knowledge in the form of class hypervectors. Its dual-mode operation enables bypassing costly feature extraction for simpler datasets, while progressive search reduces complexity by up to 61% by encoding and comparing only partial query hy"},"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":"2507.17953","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.AR","submitted_at":"2025-07-23T21:50:28Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"1b304c02a76780d5c19397f8a6209ef52bfc2939b99854d445acdf44929f1b31","abstract_canon_sha256":"6b804e15dfa91d827996d3588a51601469f06bae2b1c6fcbe8a04a9386de8817"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:42:38.077702Z","signature_b64":"ueGrQ9e2ZZwjizaEtjHUv+gldNKIvMQiewSxQbsnrbidmPp3Uwn1ZXkizpaWGkALE/6BQ/HevQ7QgkCH7WV3DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ff18c05da0fd7a7667ad2d81e18608fdd245eb2e2a118dee597d214e7e3b12a5","last_reissued_at":"2026-07-05T11:42:38.077208Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:42:38.077208Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Clo-HDnn: A 4.66 TFLOPS/W and 3.78 TOPS/W Continual On-Device Learning Accelerator with Energy-efficient Hyperdimensional Computing via Progressive Search","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AR","authors_text":"Carlos H. Diaz, Chang Eun Song, Gert Cauwenberghs, Gopabandhu Hota, Haichao Yang, Keming Fan, Kerem Akarvardar, Leo Liu, Meng-Fan Chang, Mingu Kang, Soumil Jain, Tajana Rosing, Weihong Xu","submitted_at":"2025-07-23T21:50:28Z","abstract_excerpt":"Clo-HDnn is an on-device learning (ODL) accelerator designed for emerging continual learning (CL) tasks. Clo-HDnn integrates hyperdimensional computing (HDC) along with low-cost Kronecker HD Encoder and weight clustering feature extraction (WCFE) to optimize accuracy and efficiency. Clo-HDnn adopts gradient-free CL to efficiently update and store the learned knowledge in the form of class hypervectors. Its dual-mode operation enables bypassing costly feature extraction for simpler datasets, while progressive search reduces complexity by up to 61% by encoding and comparing only partial query hy"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.17953","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/2507.17953/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":"2507.17953","created_at":"2026-07-05T11:42:38.077266+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.17953v1","created_at":"2026-07-05T11:42:38.077266+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.17953","created_at":"2026-07-05T11:42:38.077266+00:00"},{"alias_kind":"pith_short_12","alias_value":"74MMAXNA7V5H","created_at":"2026-07-05T11:42:38.077266+00:00"},{"alias_kind":"pith_short_16","alias_value":"74MMAXNA7V5HMZ5N","created_at":"2026-07-05T11:42:38.077266+00:00"},{"alias_kind":"pith_short_8","alias_value":"74MMAXNA","created_at":"2026-07-05T11:42:38.077266+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/74MMAXNA7V5HMZ5NFWA6DBQI7X","json":"https://pith.science/pith/74MMAXNA7V5HMZ5NFWA6DBQI7X.json","graph_json":"https://pith.science/api/pith-number/74MMAXNA7V5HMZ5NFWA6DBQI7X/graph.json","events_json":"https://pith.science/api/pith-number/74MMAXNA7V5HMZ5NFWA6DBQI7X/events.json","paper":"https://pith.science/paper/74MMAXNA"},"agent_actions":{"view_html":"https://pith.science/pith/74MMAXNA7V5HMZ5NFWA6DBQI7X","download_json":"https://pith.science/pith/74MMAXNA7V5HMZ5NFWA6DBQI7X.json","view_paper":"https://pith.science/paper/74MMAXNA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.17953&json=true","fetch_graph":"https://pith.science/api/pith-number/74MMAXNA7V5HMZ5NFWA6DBQI7X/graph.json","fetch_events":"https://pith.science/api/pith-number/74MMAXNA7V5HMZ5NFWA6DBQI7X/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/74MMAXNA7V5HMZ5NFWA6DBQI7X/action/timestamp_anchor","attest_storage":"https://pith.science/pith/74MMAXNA7V5HMZ5NFWA6DBQI7X/action/storage_attestation","attest_author":"https://pith.science/pith/74MMAXNA7V5HMZ5NFWA6DBQI7X/action/author_attestation","sign_citation":"https://pith.science/pith/74MMAXNA7V5HMZ5NFWA6DBQI7X/action/citation_signature","submit_replication":"https://pith.science/pith/74MMAXNA7V5HMZ5NFWA6DBQI7X/action/replication_record"}},"created_at":"2026-07-05T11:42:38.077266+00:00","updated_at":"2026-07-05T11:42:38.077266+00:00"}