{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:7RDFKCUWBFG55SBHJGCHDSKHZH","short_pith_number":"pith:7RDFKCUW","schema_version":"1.0","canonical_sha256":"fc46550a96094ddec827498471c947c9d64e25e113036351ca8a5d34218fed94","source":{"kind":"arxiv","id":"1812.00332","version":2},"attestation_state":"computed","paper":{"title":"ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Han Cai, Ligeng Zhu, Song Han","submitted_at":"2018-12-02T05:29:53Z","abstract_excerpt":"Neural architecture search (NAS) has a great impact by automatically designing effective neural network architectures. However, the prohibitive computational demand of conventional NAS algorithms (e.g. $10^4$ GPU hours) makes it difficult to \\emph{directly} search the architectures on large-scale tasks (e.g. ImageNet). Differentiable NAS can reduce the cost of GPU hours via a continuous representation of network architecture but suffers from the high GPU memory consumption issue (grow linearly w.r.t. candidate set size). As a result, they need to utilize~\\emph{proxy} tasks, such as training on"},"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":"1812.00332","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2018-12-02T05:29:53Z","cross_cats_sorted":["cs.CV","stat.ML"],"title_canon_sha256":"4b78fbe543486d8c8b9bdebe397758968ff63760c77cd74a79a8c232925ff3df","abstract_canon_sha256":"632fee2a748d3da51082933f6bf13d38d5fd6c46ae3207a5a4fca8d50a14aeff"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:52:51.401096Z","signature_b64":"YbFsK7yJHMpkR+MaoJ0smlAoi7HC4NPxTRvDFm0HvfNY6IX6F3nPXa3OJydEBOCbci3HBJ9WUodDm/QUzXkQAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fc46550a96094ddec827498471c947c9d64e25e113036351ca8a5d34218fed94","last_reissued_at":"2026-05-17T23:52:51.400296Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:52:51.400296Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Han Cai, Ligeng Zhu, Song Han","submitted_at":"2018-12-02T05:29:53Z","abstract_excerpt":"Neural architecture search (NAS) has a great impact by automatically designing effective neural network architectures. However, the prohibitive computational demand of conventional NAS algorithms (e.g. $10^4$ GPU hours) makes it difficult to \\emph{directly} search the architectures on large-scale tasks (e.g. ImageNet). Differentiable NAS can reduce the cost of GPU hours via a continuous representation of network architecture but suffers from the high GPU memory consumption issue (grow linearly w.r.t. candidate set size). As a result, they need to utilize~\\emph{proxy} tasks, such as training on"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1812.00332","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":""},"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":"1812.00332","created_at":"2026-05-17T23:52:51.400446+00:00"},{"alias_kind":"arxiv_version","alias_value":"1812.00332v2","created_at":"2026-05-17T23:52:51.400446+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1812.00332","created_at":"2026-05-17T23:52:51.400446+00:00"},{"alias_kind":"pith_short_12","alias_value":"7RDFKCUWBFG5","created_at":"2026-05-18T12:32:11.075285+00:00"},{"alias_kind":"pith_short_16","alias_value":"7RDFKCUWBFG55SBH","created_at":"2026-05-18T12:32:11.075285+00:00"},{"alias_kind":"pith_short_8","alias_value":"7RDFKCUW","created_at":"2026-05-18T12:32:11.075285+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":6,"sample":[{"citing_arxiv_id":"1906.11080","citing_title":"AGAN: Towards Automated Design of Generative Adversarial Networks","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2408.00601","citing_title":"AutoPV: Automatically Design Your Photovoltaic Power Forecasting Model","ref_index":37,"is_internal_anchor":true},{"citing_arxiv_id":"2605.21560","citing_title":"AutoMCU: Feasibility-First MCU Neural Network Customization via LLM-based Multi-Agent Systems","ref_index":38,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10989","citing_title":"SURGE: Surrogate Gradient Adaptation in Binary Neural Networks","ref_index":121,"is_internal_anchor":true},{"citing_arxiv_id":"2510.14235","citing_title":"Spiking Neural Network Architecture Search: A Survey","ref_index":161,"is_internal_anchor":true},{"citing_arxiv_id":"2605.12826","citing_title":"FRAME: Forensic Routing and Adaptive Multi-path Evidence Fusion for Image Manipulation Detection","ref_index":5,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10989","citing_title":"SURGE: Surrogate Gradient Adaptation in Binary Neural Networks","ref_index":121,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24492","citing_title":"Deployment-Aligned Low-Precision Neural Architecture Search for Spaceborne Edge AI","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23172","citing_title":"Efficient VQ-QAT and Mixed Vector/Linear quantized Neural Networks","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06714","citing_title":"Edge Deep Learning in Computer Vision and Medical Diagnostics: A Comprehensive Survey","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07378","citing_title":"Zero-Shot Neural Network Evaluation with Sample-Wise Activation Patterns","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07783","citing_title":"Chain-based Distillation for Effective Initialization of Variable-Sized Small Language Models","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14788","citing_title":"Sequence Search: Automated Sequence Design using Neural Architecture Search","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17476","citing_title":"Privatar: Scalable Privacy-preserving Multi-user VR via Secure Offloading","ref_index":116,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7RDFKCUWBFG55SBHJGCHDSKHZH","json":"https://pith.science/pith/7RDFKCUWBFG55SBHJGCHDSKHZH.json","graph_json":"https://pith.science/api/pith-number/7RDFKCUWBFG55SBHJGCHDSKHZH/graph.json","events_json":"https://pith.science/api/pith-number/7RDFKCUWBFG55SBHJGCHDSKHZH/events.json","paper":"https://pith.science/paper/7RDFKCUW"},"agent_actions":{"view_html":"https://pith.science/pith/7RDFKCUWBFG55SBHJGCHDSKHZH","download_json":"https://pith.science/pith/7RDFKCUWBFG55SBHJGCHDSKHZH.json","view_paper":"https://pith.science/paper/7RDFKCUW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1812.00332&json=true","fetch_graph":"https://pith.science/api/pith-number/7RDFKCUWBFG55SBHJGCHDSKHZH/graph.json","fetch_events":"https://pith.science/api/pith-number/7RDFKCUWBFG55SBHJGCHDSKHZH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7RDFKCUWBFG55SBHJGCHDSKHZH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7RDFKCUWBFG55SBHJGCHDSKHZH/action/storage_attestation","attest_author":"https://pith.science/pith/7RDFKCUWBFG55SBHJGCHDSKHZH/action/author_attestation","sign_citation":"https://pith.science/pith/7RDFKCUWBFG55SBHJGCHDSKHZH/action/citation_signature","submit_replication":"https://pith.science/pith/7RDFKCUWBFG55SBHJGCHDSKHZH/action/replication_record"}},"created_at":"2026-05-17T23:52:51.400446+00:00","updated_at":"2026-05-17T23:52:51.400446+00:00"}