{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:P53UNEKADVQBNVG2TTI46DRZYJ","short_pith_number":"pith:P53UNEKA","schema_version":"1.0","canonical_sha256":"7f774691401d6016d4da9cd1cf0e39c26e8dd965197c1de3ddbe9ffd3d1c3931","source":{"kind":"arxiv","id":"2006.06762","version":5},"attestation_state":"computed","paper":{"title":"Ansor: Generating High-Performance Tensor Programs for Deep Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NE","cs.PF","cs.PL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Ameer Haj-Ali, Chengfan Jia, Cody Hao Yu, Danyang Zhuo, Ion Stoica, Joseph E. Gonzalez, Jun Yang, Koushik Sen, Lianmin Zheng, Minmin Sun, Yida Wang, Zhao Wu","submitted_at":"2020-06-11T19:40:09Z","abstract_excerpt":"High-performance tensor programs are crucial to guarantee efficient execution of deep neural networks. However, obtaining performant tensor programs for different operators on various hardware platforms is notoriously challenging. Currently, deep learning systems rely on vendor-provided kernel libraries or various search strategies to get performant tensor programs. These approaches either require significant engineering effort to develop platform-specific optimization code or fall short of finding high-performance programs due to restricted search space and ineffective exploration strategy.\n "},"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":"2006.06762","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-06-11T19:40:09Z","cross_cats_sorted":["cs.NE","cs.PF","cs.PL","stat.ML"],"title_canon_sha256":"be4a6889195296151d709ee75092e1673d2642d931a4e1762ab44178ab48c4d8","abstract_canon_sha256":"2b95fcfefbab7fc0817ce5ea017f1566e161310c234c3583b474bb23c4df7635"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:00:53.949565Z","signature_b64":"mrU+T/3p/Er3Vx/Nd04+3IC6Lvlnp4VfqYfqX2ulEx1dyqbav8EtHOg746O2ZE1XvQw+qK3gkH8mPfjF8jfbCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7f774691401d6016d4da9cd1cf0e39c26e8dd965197c1de3ddbe9ffd3d1c3931","last_reissued_at":"2026-07-05T07:00:53.949037Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:00:53.949037Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Ansor: Generating High-Performance Tensor Programs for Deep Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NE","cs.PF","cs.PL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Ameer Haj-Ali, Chengfan Jia, Cody Hao Yu, Danyang Zhuo, Ion Stoica, Joseph E. Gonzalez, Jun Yang, Koushik Sen, Lianmin Zheng, Minmin Sun, Yida Wang, Zhao Wu","submitted_at":"2020-06-11T19:40:09Z","abstract_excerpt":"High-performance tensor programs are crucial to guarantee efficient execution of deep neural networks. However, obtaining performant tensor programs for different operators on various hardware platforms is notoriously challenging. Currently, deep learning systems rely on vendor-provided kernel libraries or various search strategies to get performant tensor programs. These approaches either require significant engineering effort to develop platform-specific optimization code or fall short of finding high-performance programs due to restricted search space and ineffective exploration strategy.\n "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2006.06762","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/2006.06762/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":"2006.06762","created_at":"2026-07-05T07:00:53.949095+00:00"},{"alias_kind":"arxiv_version","alias_value":"2006.06762v5","created_at":"2026-07-05T07:00:53.949095+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2006.06762","created_at":"2026-07-05T07:00:53.949095+00:00"},{"alias_kind":"pith_short_12","alias_value":"P53UNEKADVQB","created_at":"2026-07-05T07:00:53.949095+00:00"},{"alias_kind":"pith_short_16","alias_value":"P53UNEKADVQBNVG2","created_at":"2026-07-05T07:00:53.949095+00:00"},{"alias_kind":"pith_short_8","alias_value":"P53UNEKA","created_at":"2026-07-05T07:00:53.949095+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.29734","citing_title":"HTAM: Hierarchical Transition-Attended Memory for Operator Optimization","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11109","citing_title":"Record-Remix-Replay: Hierarchical GPU Kernel Optimization using Evolutionary Search","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15272","citing_title":"Prism: Symbolic Superoptimization of Tensor Programs","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2604.26039","citing_title":"RaMP: Runtime-Aware Megakernel Polymorphism for Mixture-of-Experts","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/P53UNEKADVQBNVG2TTI46DRZYJ","json":"https://pith.science/pith/P53UNEKADVQBNVG2TTI46DRZYJ.json","graph_json":"https://pith.science/api/pith-number/P53UNEKADVQBNVG2TTI46DRZYJ/graph.json","events_json":"https://pith.science/api/pith-number/P53UNEKADVQBNVG2TTI46DRZYJ/events.json","paper":"https://pith.science/paper/P53UNEKA"},"agent_actions":{"view_html":"https://pith.science/pith/P53UNEKADVQBNVG2TTI46DRZYJ","download_json":"https://pith.science/pith/P53UNEKADVQBNVG2TTI46DRZYJ.json","view_paper":"https://pith.science/paper/P53UNEKA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2006.06762&json=true","fetch_graph":"https://pith.science/api/pith-number/P53UNEKADVQBNVG2TTI46DRZYJ/graph.json","fetch_events":"https://pith.science/api/pith-number/P53UNEKADVQBNVG2TTI46DRZYJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/P53UNEKADVQBNVG2TTI46DRZYJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/P53UNEKADVQBNVG2TTI46DRZYJ/action/storage_attestation","attest_author":"https://pith.science/pith/P53UNEKADVQBNVG2TTI46DRZYJ/action/author_attestation","sign_citation":"https://pith.science/pith/P53UNEKADVQBNVG2TTI46DRZYJ/action/citation_signature","submit_replication":"https://pith.science/pith/P53UNEKADVQBNVG2TTI46DRZYJ/action/replication_record"}},"created_at":"2026-07-05T07:00:53.949095+00:00","updated_at":"2026-07-05T07:00:53.949095+00:00"}