{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:IIZ4QAS3LVHDCLVAQYWKTCQSII","short_pith_number":"pith:IIZ4QAS3","schema_version":"1.0","canonical_sha256":"4233c8025b5d4e312ea0862ca98a124225c9d64ce6b0ddf08feb2c0c0ed56212","source":{"kind":"arxiv","id":"2310.19991","version":2},"attestation_state":"computed","paper":{"title":"PolyThrottle: Energy-efficient Neural Network Inference on Edge Devices","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AR"],"primary_cat":"cs.LG","authors_text":"Hongyi Wang, Minghao Yan, Shivaram Venkataraman","submitted_at":"2023-10-30T20:19:41Z","abstract_excerpt":"As neural networks (NN) are deployed across diverse sectors, their energy demand correspondingly grows. While several prior works have focused on reducing energy consumption during training, the continuous operation of ML-powered systems leads to significant energy use during inference. This paper investigates how the configuration of on-device hardware-elements such as GPU, memory, and CPU frequency, often neglected in prior studies, affects energy consumption for NN inference with regular fine-tuning. We propose PolyThrottle, a solution that optimizes configurations across individual hardwar"},"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":"2310.19991","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-30T20:19:41Z","cross_cats_sorted":["cs.AR"],"title_canon_sha256":"7254a3521548fc40ea4c7a9ec32b01d0898ea0955e33613029be6244afe1989e","abstract_canon_sha256":"03804d320f88a01e7db4ee234133431269c9ab6a71893d67bc3782fc5a43eab9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:31:34.796433Z","signature_b64":"FpHDNiCtqekRLQl+LyPKDHcbWnKK4tmDdJ1fNEREHXjdoojGqCoPE+mFGPbchKUN0JaiDxYFLXw/RcsiWeI+DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4233c8025b5d4e312ea0862ca98a124225c9d64ce6b0ddf08feb2c0c0ed56212","last_reissued_at":"2026-07-05T07:31:34.795885Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:31:34.795885Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PolyThrottle: Energy-efficient Neural Network Inference on Edge Devices","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AR"],"primary_cat":"cs.LG","authors_text":"Hongyi Wang, Minghao Yan, Shivaram Venkataraman","submitted_at":"2023-10-30T20:19:41Z","abstract_excerpt":"As neural networks (NN) are deployed across diverse sectors, their energy demand correspondingly grows. While several prior works have focused on reducing energy consumption during training, the continuous operation of ML-powered systems leads to significant energy use during inference. This paper investigates how the configuration of on-device hardware-elements such as GPU, memory, and CPU frequency, often neglected in prior studies, affects energy consumption for NN inference with regular fine-tuning. We propose PolyThrottle, a solution that optimizes configurations across individual hardwar"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.19991","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/2310.19991/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":"2310.19991","created_at":"2026-07-05T07:31:34.795942+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.19991v2","created_at":"2026-07-05T07:31:34.795942+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.19991","created_at":"2026-07-05T07:31:34.795942+00:00"},{"alias_kind":"pith_short_12","alias_value":"IIZ4QAS3LVHD","created_at":"2026-07-05T07:31:34.795942+00:00"},{"alias_kind":"pith_short_16","alias_value":"IIZ4QAS3LVHDCLVA","created_at":"2026-07-05T07:31:34.795942+00:00"},{"alias_kind":"pith_short_8","alias_value":"IIZ4QAS3","created_at":"2026-07-05T07:31:34.795942+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.28175","citing_title":"Strait: Perceiving Priority and Interference in ML Inference Serving","ref_index":105,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10474","citing_title":"Formally Verifying Analog Neural Networks Under Process Variations Using Polynomial Zonotopes","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IIZ4QAS3LVHDCLVAQYWKTCQSII","json":"https://pith.science/pith/IIZ4QAS3LVHDCLVAQYWKTCQSII.json","graph_json":"https://pith.science/api/pith-number/IIZ4QAS3LVHDCLVAQYWKTCQSII/graph.json","events_json":"https://pith.science/api/pith-number/IIZ4QAS3LVHDCLVAQYWKTCQSII/events.json","paper":"https://pith.science/paper/IIZ4QAS3"},"agent_actions":{"view_html":"https://pith.science/pith/IIZ4QAS3LVHDCLVAQYWKTCQSII","download_json":"https://pith.science/pith/IIZ4QAS3LVHDCLVAQYWKTCQSII.json","view_paper":"https://pith.science/paper/IIZ4QAS3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.19991&json=true","fetch_graph":"https://pith.science/api/pith-number/IIZ4QAS3LVHDCLVAQYWKTCQSII/graph.json","fetch_events":"https://pith.science/api/pith-number/IIZ4QAS3LVHDCLVAQYWKTCQSII/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IIZ4QAS3LVHDCLVAQYWKTCQSII/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IIZ4QAS3LVHDCLVAQYWKTCQSII/action/storage_attestation","attest_author":"https://pith.science/pith/IIZ4QAS3LVHDCLVAQYWKTCQSII/action/author_attestation","sign_citation":"https://pith.science/pith/IIZ4QAS3LVHDCLVAQYWKTCQSII/action/citation_signature","submit_replication":"https://pith.science/pith/IIZ4QAS3LVHDCLVAQYWKTCQSII/action/replication_record"}},"created_at":"2026-07-05T07:31:34.795942+00:00","updated_at":"2026-07-05T07:31:34.795942+00:00"}