{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:6UFGCKNC45HA47UBWKHAMPSMBS","short_pith_number":"pith:6UFGCKNC","schema_version":"1.0","canonical_sha256":"f50a6129a2e74e0e7e81b28e063e4c0c8ac1124dffd2715ca820ed7a30c99d3f","source":{"kind":"arxiv","id":"2205.10034","version":3},"attestation_state":"computed","paper":{"title":"MoESys: A Distributed and Efficient Mixture-of-Experts Training and Inference System for Internet Services","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.DC","authors_text":"Dianhai Yu, Haoyi Xiong, Hongxiang Hao, Huachao Wu, Jiang Bian, Liang Shen, Lirong Dai, Weibao Gong","submitted_at":"2022-05-20T09:09:27Z","abstract_excerpt":"While modern internet services, such as chatbots, search engines, and online advertising, demand the use of large-scale deep neural networks (DNNs), distributed training and inference over heterogeneous computing systems are desired to facilitate these DNN models. Mixture-of-Experts (MoE) is one the most common strategies to lower the cost of training subject to the overall size of models/data through gating and parallelism in a divide-and-conquer fashion. While DeepSpeed has made efforts in carrying out large-scale MoE training over heterogeneous infrastructures, the efficiency of training an"},"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":"2205.10034","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2022-05-20T09:09:27Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"e46848a91facb74f21dcaf64542196e21df306284817c8341e872f0de81096e0","abstract_canon_sha256":"72365ddbc6f8c9fb50ca559f742627d930b303446fcbd3b86aa88ea3ce100353"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:54:00.168876Z","signature_b64":"nSF3XfF9jWJFwHJYT+j95BEB0FQUvke7DYms58gD5DNVDDfzGZpSq9a42dovEeD3vbxjduds0DcGC4I+f4bTAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f50a6129a2e74e0e7e81b28e063e4c0c8ac1124dffd2715ca820ed7a30c99d3f","last_reissued_at":"2026-07-05T08:54:00.168464Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:54:00.168464Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MoESys: A Distributed and Efficient Mixture-of-Experts Training and Inference System for Internet Services","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.DC","authors_text":"Dianhai Yu, Haoyi Xiong, Hongxiang Hao, Huachao Wu, Jiang Bian, Liang Shen, Lirong Dai, Weibao Gong","submitted_at":"2022-05-20T09:09:27Z","abstract_excerpt":"While modern internet services, such as chatbots, search engines, and online advertising, demand the use of large-scale deep neural networks (DNNs), distributed training and inference over heterogeneous computing systems are desired to facilitate these DNN models. Mixture-of-Experts (MoE) is one the most common strategies to lower the cost of training subject to the overall size of models/data through gating and parallelism in a divide-and-conquer fashion. While DeepSpeed has made efforts in carrying out large-scale MoE training over heterogeneous infrastructures, the efficiency of training an"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.10034","kind":"arxiv","version":3},"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/2205.10034/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":"2205.10034","created_at":"2026-07-05T08:54:00.168523+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.10034v3","created_at":"2026-07-05T08:54:00.168523+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.10034","created_at":"2026-07-05T08:54:00.168523+00:00"},{"alias_kind":"pith_short_12","alias_value":"6UFGCKNC45HA","created_at":"2026-07-05T08:54:00.168523+00:00"},{"alias_kind":"pith_short_16","alias_value":"6UFGCKNC45HA47UB","created_at":"2026-07-05T08:54:00.168523+00:00"},{"alias_kind":"pith_short_8","alias_value":"6UFGCKNC","created_at":"2026-07-05T08:54:00.168523+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.23266","citing_title":"Sub-MoE: Efficient Mixture-of-Expert LLMs Compression via Subspace Expert Merging","ref_index":34,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6UFGCKNC45HA47UBWKHAMPSMBS","json":"https://pith.science/pith/6UFGCKNC45HA47UBWKHAMPSMBS.json","graph_json":"https://pith.science/api/pith-number/6UFGCKNC45HA47UBWKHAMPSMBS/graph.json","events_json":"https://pith.science/api/pith-number/6UFGCKNC45HA47UBWKHAMPSMBS/events.json","paper":"https://pith.science/paper/6UFGCKNC"},"agent_actions":{"view_html":"https://pith.science/pith/6UFGCKNC45HA47UBWKHAMPSMBS","download_json":"https://pith.science/pith/6UFGCKNC45HA47UBWKHAMPSMBS.json","view_paper":"https://pith.science/paper/6UFGCKNC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.10034&json=true","fetch_graph":"https://pith.science/api/pith-number/6UFGCKNC45HA47UBWKHAMPSMBS/graph.json","fetch_events":"https://pith.science/api/pith-number/6UFGCKNC45HA47UBWKHAMPSMBS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6UFGCKNC45HA47UBWKHAMPSMBS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6UFGCKNC45HA47UBWKHAMPSMBS/action/storage_attestation","attest_author":"https://pith.science/pith/6UFGCKNC45HA47UBWKHAMPSMBS/action/author_attestation","sign_citation":"https://pith.science/pith/6UFGCKNC45HA47UBWKHAMPSMBS/action/citation_signature","submit_replication":"https://pith.science/pith/6UFGCKNC45HA47UBWKHAMPSMBS/action/replication_record"}},"created_at":"2026-07-05T08:54:00.168523+00:00","updated_at":"2026-07-05T08:54:00.168523+00:00"}