{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:K45YU5Y2FTQXLIAHVB2NVU3TQY","short_pith_number":"pith:K45YU5Y2","schema_version":"1.0","canonical_sha256":"573b8a771a2ce175a007a874dad373862544493f7a438f1805835512d4e083ef","source":{"kind":"arxiv","id":"2405.17976","version":2},"attestation_state":"computed","paper":{"title":"Yuan 2.0-M32: Mixture of Experts with Attention Router","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Chao Wang, Chong Shen, Fei Wang, Houbo He, Jiangang Luo, Junxiong Mao, Lingjun Li, Shaohua Wu, Tong Yu, Weixu Qiao, Xi Chen, Xudong Zhao, Yue Wang, Zeru Zhang, Zeyu Sun","submitted_at":"2024-05-28T09:05:08Z","abstract_excerpt":"Yuan 2.0-M32, with a similar base architecture as Yuan-2.0 2B, uses a mixture-of-experts architecture with 32 experts of which 2 experts are active. A new router network, Attention Router, is proposed and adopted for a more efficient selection of experts, which improves the accuracy compared to the model with classical router network. Yuan 2.0-M32 is trained with 2000B tokens from scratch, and the training computation consumption is only 9.25% of a dense model at the same parameter scale. Yuan 2.0-M32 demonstrates competitive capability on coding, math, and various domains of expertise, with o"},"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":"2405.17976","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-05-28T09:05:08Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"b156f7fad35ee8605f12b23208f90fb5575fe9f52a84d9a3925cce45988bc9f6","abstract_canon_sha256":"f984b8207f23e4694f8af087d12e293b6c91f4efad5c34479e6d51dea94133b7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:24:25.576382Z","signature_b64":"/LT3u7S+CjqvRjDZaRHKYhAtvIeHPHtqWFBH96L3tg7hz322/Qu4vq9XI0FbVXcBIQ5daVlMP+JV6t3HcbQOCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"573b8a771a2ce175a007a874dad373862544493f7a438f1805835512d4e083ef","last_reissued_at":"2026-07-05T08:24:25.575824Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:24:25.575824Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Yuan 2.0-M32: Mixture of Experts with Attention Router","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Chao Wang, Chong Shen, Fei Wang, Houbo He, Jiangang Luo, Junxiong Mao, Lingjun Li, Shaohua Wu, Tong Yu, Weixu Qiao, Xi Chen, Xudong Zhao, Yue Wang, Zeru Zhang, Zeyu Sun","submitted_at":"2024-05-28T09:05:08Z","abstract_excerpt":"Yuan 2.0-M32, with a similar base architecture as Yuan-2.0 2B, uses a mixture-of-experts architecture with 32 experts of which 2 experts are active. A new router network, Attention Router, is proposed and adopted for a more efficient selection of experts, which improves the accuracy compared to the model with classical router network. Yuan 2.0-M32 is trained with 2000B tokens from scratch, and the training computation consumption is only 9.25% of a dense model at the same parameter scale. Yuan 2.0-M32 demonstrates competitive capability on coding, math, and various domains of expertise, with o"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.17976","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/2405.17976/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":"2405.17976","created_at":"2026-07-05T08:24:25.575896+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.17976v2","created_at":"2026-07-05T08:24:25.575896+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.17976","created_at":"2026-07-05T08:24:25.575896+00:00"},{"alias_kind":"pith_short_12","alias_value":"K45YU5Y2FTQX","created_at":"2026-07-05T08:24:25.575896+00:00"},{"alias_kind":"pith_short_16","alias_value":"K45YU5Y2FTQXLIAH","created_at":"2026-07-05T08:24:25.575896+00:00"},{"alias_kind":"pith_short_8","alias_value":"K45YU5Y2","created_at":"2026-07-05T08:24:25.575896+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.12119","citing_title":"Mixture-of-Experts Can Surpass Dense LLMs Under Strictly Equal Resource","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2601.21349","citing_title":"L2R: Low-Rank and Lipschitz-Controlled Routing for Mixture-of-Experts","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/K45YU5Y2FTQXLIAHVB2NVU3TQY","json":"https://pith.science/pith/K45YU5Y2FTQXLIAHVB2NVU3TQY.json","graph_json":"https://pith.science/api/pith-number/K45YU5Y2FTQXLIAHVB2NVU3TQY/graph.json","events_json":"https://pith.science/api/pith-number/K45YU5Y2FTQXLIAHVB2NVU3TQY/events.json","paper":"https://pith.science/paper/K45YU5Y2"},"agent_actions":{"view_html":"https://pith.science/pith/K45YU5Y2FTQXLIAHVB2NVU3TQY","download_json":"https://pith.science/pith/K45YU5Y2FTQXLIAHVB2NVU3TQY.json","view_paper":"https://pith.science/paper/K45YU5Y2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.17976&json=true","fetch_graph":"https://pith.science/api/pith-number/K45YU5Y2FTQXLIAHVB2NVU3TQY/graph.json","fetch_events":"https://pith.science/api/pith-number/K45YU5Y2FTQXLIAHVB2NVU3TQY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/K45YU5Y2FTQXLIAHVB2NVU3TQY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/K45YU5Y2FTQXLIAHVB2NVU3TQY/action/storage_attestation","attest_author":"https://pith.science/pith/K45YU5Y2FTQXLIAHVB2NVU3TQY/action/author_attestation","sign_citation":"https://pith.science/pith/K45YU5Y2FTQXLIAHVB2NVU3TQY/action/citation_signature","submit_replication":"https://pith.science/pith/K45YU5Y2FTQXLIAHVB2NVU3TQY/action/replication_record"}},"created_at":"2026-07-05T08:24:25.575896+00:00","updated_at":"2026-07-05T08:24:25.575896+00:00"}