{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:JQSXXCPUE367ZCZGZP5HEA2V74","short_pith_number":"pith:JQSXXCPU","schema_version":"1.0","canonical_sha256":"4c257b89f426fdfc8b26cbfa720355ff065fd7189b6adcf96ca19eb2d75cd919","source":{"kind":"arxiv","id":"2206.00277","version":2},"attestation_state":"computed","paper":{"title":"Task-Specific Expert Pruning for Sparse Mixture-of-Experts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Binxing Jiao, Daxin Jiang, Furu Wei, Haoyi Zhou, Jianxin Li, Shaohan Huang, Tianyu Chen, Yuan Xie","submitted_at":"2022-06-01T07:09:01Z","abstract_excerpt":"The sparse Mixture-of-Experts (MoE) model is powerful for large-scale pre-training and has achieved promising results due to its model capacity. However, with trillions of parameters, MoE is hard to be deployed on cloud or mobile environment. The inference of MoE requires expert parallelism, which is not hardware-friendly and communication expensive. Especially for resource-limited downstream tasks, such sparse structure has to sacrifice a lot of computing efficiency for limited performance gains. In this work, we observe most experts contribute scarcely little to the MoE fine-tuning and infer"},"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":"2206.00277","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-06-01T07:09:01Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2c86d4fb47418bd4ad58e8200740d1434d440ee1ca909c4c4904c90ebc9e8781","abstract_canon_sha256":"99933bb70fe5a21b95d381a211623dd19b90302a3ff60e9672b6b8eace243cdf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:28:32.485167Z","signature_b64":"BFsTl2HxomIdgcAjLJ05gsEwJihLJlpzUOii/E/ckPT1hv6FNicVQsg5Ebnooau6IBzYX2GrHSctovSkjnY1Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4c257b89f426fdfc8b26cbfa720355ff065fd7189b6adcf96ca19eb2d75cd919","last_reissued_at":"2026-07-05T04:28:32.484684Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:28:32.484684Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Task-Specific Expert Pruning for Sparse Mixture-of-Experts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Binxing Jiao, Daxin Jiang, Furu Wei, Haoyi Zhou, Jianxin Li, Shaohan Huang, Tianyu Chen, Yuan Xie","submitted_at":"2022-06-01T07:09:01Z","abstract_excerpt":"The sparse Mixture-of-Experts (MoE) model is powerful for large-scale pre-training and has achieved promising results due to its model capacity. However, with trillions of parameters, MoE is hard to be deployed on cloud or mobile environment. The inference of MoE requires expert parallelism, which is not hardware-friendly and communication expensive. Especially for resource-limited downstream tasks, such sparse structure has to sacrifice a lot of computing efficiency for limited performance gains. In this work, we observe most experts contribute scarcely little to the MoE fine-tuning and infer"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.00277","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/2206.00277/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":"2206.00277","created_at":"2026-07-05T04:28:32.484743+00:00"},{"alias_kind":"arxiv_version","alias_value":"2206.00277v2","created_at":"2026-07-05T04:28:32.484743+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.00277","created_at":"2026-07-05T04:28:32.484743+00:00"},{"alias_kind":"pith_short_12","alias_value":"JQSXXCPUE367","created_at":"2026-07-05T04:28:32.484743+00:00"},{"alias_kind":"pith_short_16","alias_value":"JQSXXCPUE367ZCZG","created_at":"2026-07-05T04:28:32.484743+00:00"},{"alias_kind":"pith_short_8","alias_value":"JQSXXCPU","created_at":"2026-07-05T04:28:32.484743+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.01444","citing_title":"On the Utility and Factual Reliability of Pruned Mixture-of-Experts Models in the Biomedical Domain","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05538","citing_title":"Less is MoE: Trimming Experts in Domain-Specialist Language Models","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09886","citing_title":"SHAPE: Coalition-Aware Expert Pruning for Sparse Mixture-of-Experts LLMs","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01007","citing_title":"Beyond Task-Agnostic: Task-Aware Grouping for Communication-Efficient Multi-Task MoE Inference","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29982","citing_title":"Beyond Uniform Experts: Cost-Aware Expert Execution for Efficient Multi-Device MoE Inference","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28207","citing_title":"Pruning and Distilling Mixture-of-Experts into Dense Language Models","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30876","citing_title":"dMoE: dLLMs with Learnable Block Experts","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2411.08982","citing_title":"Lynx: Enabling Efficient MoE Inference through Dynamic Batch-Aware Expert Selection","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2506.12876","citing_title":"MaskPro: Linear-Space Probabilistic Learning for Strict (N:M)-Sparsity on LLMs","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2509.07379","citing_title":"DuoServe-MoE: Dual-Phase Expert Prefetch and Caching for LLM Inference QoS Assurance","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2603.06003","citing_title":"EvoESAP: Non-Uniform Expert Pruning for Sparse MoE","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02715","citing_title":"FluxMoE: Decoupling Expert Residency for High-Performance MoE Serving","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06542","citing_title":"Does a Global Perspective Help Prune Sparse MoEs Elegantly?","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04356","citing_title":"REAM: Merging Improves Pruning of Experts in LLMs","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JQSXXCPUE367ZCZGZP5HEA2V74","json":"https://pith.science/pith/JQSXXCPUE367ZCZGZP5HEA2V74.json","graph_json":"https://pith.science/api/pith-number/JQSXXCPUE367ZCZGZP5HEA2V74/graph.json","events_json":"https://pith.science/api/pith-number/JQSXXCPUE367ZCZGZP5HEA2V74/events.json","paper":"https://pith.science/paper/JQSXXCPU"},"agent_actions":{"view_html":"https://pith.science/pith/JQSXXCPUE367ZCZGZP5HEA2V74","download_json":"https://pith.science/pith/JQSXXCPUE367ZCZGZP5HEA2V74.json","view_paper":"https://pith.science/paper/JQSXXCPU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2206.00277&json=true","fetch_graph":"https://pith.science/api/pith-number/JQSXXCPUE367ZCZGZP5HEA2V74/graph.json","fetch_events":"https://pith.science/api/pith-number/JQSXXCPUE367ZCZGZP5HEA2V74/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JQSXXCPUE367ZCZGZP5HEA2V74/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JQSXXCPUE367ZCZGZP5HEA2V74/action/storage_attestation","attest_author":"https://pith.science/pith/JQSXXCPUE367ZCZGZP5HEA2V74/action/author_attestation","sign_citation":"https://pith.science/pith/JQSXXCPUE367ZCZGZP5HEA2V74/action/citation_signature","submit_replication":"https://pith.science/pith/JQSXXCPUE367ZCZGZP5HEA2V74/action/replication_record"}},"created_at":"2026-07-05T04:28:32.484743+00:00","updated_at":"2026-07-05T04:28:32.484743+00:00"}