{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NYYM5AKOURDM62TBFUJDIAOG2E","short_pith_number":"pith:NYYM5AKO","schema_version":"1.0","canonical_sha256":"6e30ce814ea446cf6a612d123401c6d13da67a0222bddce5076ccc62a99bdfb1","source":{"kind":"arxiv","id":"2508.12801","version":1},"attestation_state":"computed","paper":{"title":"Maximum Score Routing For Mixture-of-Experts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Bowen Dong, Jianyong Wang, Tengyu Pan, Xun Zhou, Yilong Fan, Yutao Sun, Zhenyu Li","submitted_at":"2025-08-18T10:25:42Z","abstract_excerpt":"Routing networks in sparsely activated mixture-of-experts (MoE) dynamically allocate input tokens to top-k experts through differentiable sparse transformations, enabling scalable model capacity while preserving computational efficiency. Traditional MoE networks impose an expert capacity constraint to ensure GPU-friendly computation. However, this leads to token dropping when capacity is saturated and results in low hardware efficiency due to padding in underutilized experts. Removing the capacity constraint, in turn, compromises load balancing and computational efficiency. To address these is"},"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":"2508.12801","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-08-18T10:25:42Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"5625a6defd133f8689177ad542a96451f39e7f45dac607019c954922aa3762b8","abstract_canon_sha256":"c75208f5114ef38778e6ea708520c317f6ab8d36517b0b6db963bdfcaebfeb2c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:55:27.444454Z","signature_b64":"9apcKUWqvgJQaPEfLE6ZHTdZ0nR5JOKrGoUz7lR2v4M7IdrhklkALWwWAy+OMcDjJ+5ej5NSmVA61mP0Y6wIDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6e30ce814ea446cf6a612d123401c6d13da67a0222bddce5076ccc62a99bdfb1","last_reissued_at":"2026-07-05T11:55:27.443940Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:55:27.443940Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Maximum Score Routing For Mixture-of-Experts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Bowen Dong, Jianyong Wang, Tengyu Pan, Xun Zhou, Yilong Fan, Yutao Sun, Zhenyu Li","submitted_at":"2025-08-18T10:25:42Z","abstract_excerpt":"Routing networks in sparsely activated mixture-of-experts (MoE) dynamically allocate input tokens to top-k experts through differentiable sparse transformations, enabling scalable model capacity while preserving computational efficiency. Traditional MoE networks impose an expert capacity constraint to ensure GPU-friendly computation. However, this leads to token dropping when capacity is saturated and results in low hardware efficiency due to padding in underutilized experts. Removing the capacity constraint, in turn, compromises load balancing and computational efficiency. To address these is"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.12801","kind":"arxiv","version":1},"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/2508.12801/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":"2508.12801","created_at":"2026-07-05T11:55:27.444000+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.12801v1","created_at":"2026-07-05T11:55:27.444000+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.12801","created_at":"2026-07-05T11:55:27.444000+00:00"},{"alias_kind":"pith_short_12","alias_value":"NYYM5AKOURDM","created_at":"2026-07-05T11:55:27.444000+00:00"},{"alias_kind":"pith_short_16","alias_value":"NYYM5AKOURDM62TB","created_at":"2026-07-05T11:55:27.444000+00:00"},{"alias_kind":"pith_short_8","alias_value":"NYYM5AKO","created_at":"2026-07-05T11:55:27.444000+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/NYYM5AKOURDM62TBFUJDIAOG2E","json":"https://pith.science/pith/NYYM5AKOURDM62TBFUJDIAOG2E.json","graph_json":"https://pith.science/api/pith-number/NYYM5AKOURDM62TBFUJDIAOG2E/graph.json","events_json":"https://pith.science/api/pith-number/NYYM5AKOURDM62TBFUJDIAOG2E/events.json","paper":"https://pith.science/paper/NYYM5AKO"},"agent_actions":{"view_html":"https://pith.science/pith/NYYM5AKOURDM62TBFUJDIAOG2E","download_json":"https://pith.science/pith/NYYM5AKOURDM62TBFUJDIAOG2E.json","view_paper":"https://pith.science/paper/NYYM5AKO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.12801&json=true","fetch_graph":"https://pith.science/api/pith-number/NYYM5AKOURDM62TBFUJDIAOG2E/graph.json","fetch_events":"https://pith.science/api/pith-number/NYYM5AKOURDM62TBFUJDIAOG2E/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NYYM5AKOURDM62TBFUJDIAOG2E/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NYYM5AKOURDM62TBFUJDIAOG2E/action/storage_attestation","attest_author":"https://pith.science/pith/NYYM5AKOURDM62TBFUJDIAOG2E/action/author_attestation","sign_citation":"https://pith.science/pith/NYYM5AKOURDM62TBFUJDIAOG2E/action/citation_signature","submit_replication":"https://pith.science/pith/NYYM5AKOURDM62TBFUJDIAOG2E/action/replication_record"}},"created_at":"2026-07-05T11:55:27.444000+00:00","updated_at":"2026-07-05T11:55:27.444000+00:00"}