{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:THGSX5X43OECVHC3RLEDMXPMAX","short_pith_number":"pith:THGSX5X4","schema_version":"1.0","canonical_sha256":"99cd2bf6fcdb882a9c5b8ac8365dec05f0a86025ac71bc8d38b96c088755bb2a","source":{"kind":"arxiv","id":"2202.09368","version":2},"attestation_state":"computed","paper":{"title":"Mixture-of-Experts with Expert Choice Routing","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Andrew Dai, Hanxiao Liu, James Laudon, Nan Du, Quoc Le, Tao Lei, Vincent Zhao, Yanping Huang, Yanqi Zhou, Zhifeng Chen","submitted_at":"2022-02-18T17:46:11Z","abstract_excerpt":"Sparsely-activated Mixture-of-experts (MoE) models allow the number of parameters to greatly increase while keeping the amount of computation for a given token or a given sample unchanged. However, a poor expert routing strategy (e.g. one resulting in load imbalance) can cause certain experts to be under-trained, leading to an expert being under or over-specialized. Prior work allocates a fixed number of experts to each token using a top-k function regardless of the relative importance of different tokens. To address this, we propose a heterogeneous mixture-of-experts employing an expert choic"},"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":"2202.09368","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2022-02-18T17:46:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8e334ca010148e529275f8f3a5999058d6afc162aaac1b8ded9bd48c66b10179","abstract_canon_sha256":"5d762f2e07ddf533b654ba3b5c8c6e5c25520f0211072547c2594baae9585060"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:06:30.199993Z","signature_b64":"mtH1cDQUGciXCOUjuJ3+B5Q5n+x1pu7shb2cVKJm6lRa6gRavm4TAhEAjZjM/O/KiIEZdE6H2/tG1Nab2aKFDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"99cd2bf6fcdb882a9c5b8ac8365dec05f0a86025ac71bc8d38b96c088755bb2a","last_reissued_at":"2026-07-05T05:06:30.199582Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:06:30.199582Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mixture-of-Experts with Expert Choice Routing","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Andrew Dai, Hanxiao Liu, James Laudon, Nan Du, Quoc Le, Tao Lei, Vincent Zhao, Yanping Huang, Yanqi Zhou, Zhifeng Chen","submitted_at":"2022-02-18T17:46:11Z","abstract_excerpt":"Sparsely-activated Mixture-of-experts (MoE) models allow the number of parameters to greatly increase while keeping the amount of computation for a given token or a given sample unchanged. However, a poor expert routing strategy (e.g. one resulting in load imbalance) can cause certain experts to be under-trained, leading to an expert being under or over-specialized. Prior work allocates a fixed number of experts to each token using a top-k function regardless of the relative importance of different tokens. To address this, we propose a heterogeneous mixture-of-experts employing an expert choic"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2202.09368","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/2202.09368/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":"2202.09368","created_at":"2026-07-05T05:06:30.199635+00:00"},{"alias_kind":"arxiv_version","alias_value":"2202.09368v2","created_at":"2026-07-05T05:06:30.199635+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2202.09368","created_at":"2026-07-05T05:06:30.199635+00:00"},{"alias_kind":"pith_short_12","alias_value":"THGSX5X43OEC","created_at":"2026-07-05T05:06:30.199635+00:00"},{"alias_kind":"pith_short_16","alias_value":"THGSX5X43OECVHC3","created_at":"2026-07-05T05:06:30.199635+00:00"},{"alias_kind":"pith_short_8","alias_value":"THGSX5X4","created_at":"2026-07-05T05:06:30.199635+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":11,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.31363","citing_title":"Language-Assisted Super-Resolution from Real-World Low-Resolution Patches","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2607.01710","citing_title":"Generic Expert Coverage for Pruning SparseMixture-of-Experts Language Models","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01143","citing_title":"Schedule-Level Shared-Prefix Reuse for LLM RL Training","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31363","citing_title":"Language-Assisted Super-Resolution from Real-World Low-Resolution Patches","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29121","citing_title":"A Minimal Bifurcation Model of Load Imbalance in a Softmax Mixture-of-Experts Router","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2411.04996","citing_title":"Mixture-of-Transformers: A Sparse and Scalable Architecture for Multi-Modal Foundation Models","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2408.15664","citing_title":"Auxiliary-Loss-Free Load Balancing Strategy for Mixture-of-Experts","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00539","citing_title":"AGoQ: Activation and Gradient Quantization for Memory-Efficient Distributed Training of LLMs","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00604","citing_title":"Affinity Is Not Enough: Recovering the Free Energy Principle in Mixture-of-Experts","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00539","citing_title":"AGoQ: Activation and Gradient Quantization for Memory-Efficient Distributed Training of LLMs","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05049","citing_title":"Piper: Efficient Large-Scale MoE Training via Resource Modeling and Pipelined Hybrid Parallelism","ref_index":32,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/THGSX5X43OECVHC3RLEDMXPMAX","json":"https://pith.science/pith/THGSX5X43OECVHC3RLEDMXPMAX.json","graph_json":"https://pith.science/api/pith-number/THGSX5X43OECVHC3RLEDMXPMAX/graph.json","events_json":"https://pith.science/api/pith-number/THGSX5X43OECVHC3RLEDMXPMAX/events.json","paper":"https://pith.science/paper/THGSX5X4"},"agent_actions":{"view_html":"https://pith.science/pith/THGSX5X43OECVHC3RLEDMXPMAX","download_json":"https://pith.science/pith/THGSX5X43OECVHC3RLEDMXPMAX.json","view_paper":"https://pith.science/paper/THGSX5X4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2202.09368&json=true","fetch_graph":"https://pith.science/api/pith-number/THGSX5X43OECVHC3RLEDMXPMAX/graph.json","fetch_events":"https://pith.science/api/pith-number/THGSX5X43OECVHC3RLEDMXPMAX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/THGSX5X43OECVHC3RLEDMXPMAX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/THGSX5X43OECVHC3RLEDMXPMAX/action/storage_attestation","attest_author":"https://pith.science/pith/THGSX5X43OECVHC3RLEDMXPMAX/action/author_attestation","sign_citation":"https://pith.science/pith/THGSX5X43OECVHC3RLEDMXPMAX/action/citation_signature","submit_replication":"https://pith.science/pith/THGSX5X43OECVHC3RLEDMXPMAX/action/replication_record"}},"created_at":"2026-07-05T05:06:30.199635+00:00","updated_at":"2026-07-05T05:06:30.199635+00:00"}