{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:AZ66R6USWMODF2656UJ3LLJETN","short_pith_number":"pith:AZ66R6US","schema_version":"1.0","canonical_sha256":"067de8fa92b31c32ebddf513b5ad249b6bfe0ef12c2d32ddaa2ae1775248fabb","source":{"kind":"arxiv","id":"2602.05711","version":2},"attestation_state":"computed","paper":{"title":"OmniMoE: An Efficient MoE by Orchestrating Atomic Experts at Scale","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Guang Liu, Jingze Shi, Yifan Wu, Yizhang Zhu, Yuyu Luo, Zhangyang Peng","submitted_at":"2026-02-05T14:37:32Z","abstract_excerpt":"Mixture-of-Experts (MoE) architectures are evolving towards finer granularity to improve parameter efficiency. However, existing MoE designs face an inherent trade-off between the granularity of expert specialization and hardware execution efficiency. We propose OmniMoE, a system-algorithm co-designed framework that pushes expert granularity to its logical extreme. OmniMoE introduces vector-level Atomic Experts, enabling scalable routing and execution within a single MoE layer, while retaining a shared dense MLP branch for general-purpose processing. Although this atomic design maximizes capac"},"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":"2602.05711","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-05T14:37:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"20fe187d35b832497782c4adcc8e4e603b00ac66b3e445e160647ce239474997","abstract_canon_sha256":"9f72b3fdd6c47797a7edd231a34210d9dcc62fa7d7b3466e4327ac32b10c1c04"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-02T01:17:27.710734Z","signature_b64":"dt2V/DZwLh6nOzymAZvUD+tN3Z1SkC59Cse4b7UHpqUzKe8Kk7jiVok5flRUq6KO7nwgGil8bC81MKZSTPPADA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"067de8fa92b31c32ebddf513b5ad249b6bfe0ef12c2d32ddaa2ae1775248fabb","last_reissued_at":"2026-07-02T01:17:27.710293Z","signature_status":"signed_v1","first_computed_at":"2026-07-02T01:17:27.710293Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"OmniMoE: An Efficient MoE by Orchestrating Atomic Experts at Scale","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Guang Liu, Jingze Shi, Yifan Wu, Yizhang Zhu, Yuyu Luo, Zhangyang Peng","submitted_at":"2026-02-05T14:37:32Z","abstract_excerpt":"Mixture-of-Experts (MoE) architectures are evolving towards finer granularity to improve parameter efficiency. However, existing MoE designs face an inherent trade-off between the granularity of expert specialization and hardware execution efficiency. We propose OmniMoE, a system-algorithm co-designed framework that pushes expert granularity to its logical extreme. OmniMoE introduces vector-level Atomic Experts, enabling scalable routing and execution within a single MoE layer, while retaining a shared dense MLP branch for general-purpose processing. Although this atomic design maximizes capac"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.05711","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/2602.05711/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":"2602.05711","created_at":"2026-07-02T01:17:27.710365+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.05711v2","created_at":"2026-07-02T01:17:27.710365+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.05711","created_at":"2026-07-02T01:17:27.710365+00:00"},{"alias_kind":"pith_short_12","alias_value":"AZ66R6USWMOD","created_at":"2026-07-02T01:17:27.710365+00:00"},{"alias_kind":"pith_short_16","alias_value":"AZ66R6USWMODF265","created_at":"2026-07-02T01:17:27.710365+00:00"},{"alias_kind":"pith_short_8","alias_value":"AZ66R6US","created_at":"2026-07-02T01:17:27.710365+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AZ66R6USWMODF2656UJ3LLJETN","json":"https://pith.science/pith/AZ66R6USWMODF2656UJ3LLJETN.json","graph_json":"https://pith.science/api/pith-number/AZ66R6USWMODF2656UJ3LLJETN/graph.json","events_json":"https://pith.science/api/pith-number/AZ66R6USWMODF2656UJ3LLJETN/events.json","paper":"https://pith.science/paper/AZ66R6US"},"agent_actions":{"view_html":"https://pith.science/pith/AZ66R6USWMODF2656UJ3LLJETN","download_json":"https://pith.science/pith/AZ66R6USWMODF2656UJ3LLJETN.json","view_paper":"https://pith.science/paper/AZ66R6US","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.05711&json=true","fetch_graph":"https://pith.science/api/pith-number/AZ66R6USWMODF2656UJ3LLJETN/graph.json","fetch_events":"https://pith.science/api/pith-number/AZ66R6USWMODF2656UJ3LLJETN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AZ66R6USWMODF2656UJ3LLJETN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AZ66R6USWMODF2656UJ3LLJETN/action/storage_attestation","attest_author":"https://pith.science/pith/AZ66R6USWMODF2656UJ3LLJETN/action/author_attestation","sign_citation":"https://pith.science/pith/AZ66R6USWMODF2656UJ3LLJETN/action/citation_signature","submit_replication":"https://pith.science/pith/AZ66R6USWMODF2656UJ3LLJETN/action/replication_record"}},"created_at":"2026-07-02T01:17:27.710365+00:00","updated_at":"2026-07-02T01:17:27.710365+00:00"}