{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:NE7W423ILRZWFNV6RN6FMRIGIU","short_pith_number":"pith:NE7W423I","schema_version":"1.0","canonical_sha256":"693f6e6b685c7362b6be8b7c56450645135e367ede4c249cabd997f655aa4203","source":{"kind":"arxiv","id":"2605.27081","version":1},"attestation_state":"computed","paper":{"title":"ReMoE: Boosting Expert Reuse through Router Fine-Tuning in Memory-Constrained MoE LLM Inference","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.DC"],"primary_cat":"cs.LG","authors_text":"Liang Wang, Limin Xiao, Tianyang Jiang, Xiaojian Liao, Xiongwei Zhu, Yusen Zhang","submitted_at":"2026-05-26T14:32:56Z","abstract_excerpt":"Fine-grained Mixture-of-Experts (MoE) models sparsely activate only a subset of experts per token, reducing activated computation while maintaining high model capacity. However, in memory-constrained inference scenarios, only a small set of experts can be cached. Experts not in the cache must be fetched from slow external storage (e.g., UFS), leading to frequent evictions and substantial I/O overhead.\n  We propose ReMoE, a router fine-tuning framework designed to boost token-wise expert reuse. ReMoE biases the router toward recently selected experts, producing temporally stable routing that be"},"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":"2605.27081","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-26T14:32:56Z","cross_cats_sorted":["cs.AI","cs.DC"],"title_canon_sha256":"93117bad2bdca10e1a25d72d608b7b594cb619ec7604118ae9cdf04c6d94b608","abstract_canon_sha256":"0076d214a7fc81e9dac5dfedd3ea75272281e3959a28b09cd63e02dd5896f524"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T01:06:27.309432Z","signature_b64":"2VeugPQFRz3mxzdEn6TnjIPtxCt4NR/LgN+S1Zi/KzEGZ7DTcgEnWvRiIPFTH2ffUdxsxOLQuw4psKGp3TtsBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"693f6e6b685c7362b6be8b7c56450645135e367ede4c249cabd997f655aa4203","last_reissued_at":"2026-05-27T01:06:27.308937Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T01:06:27.308937Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ReMoE: Boosting Expert Reuse through Router Fine-Tuning in Memory-Constrained MoE LLM Inference","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.DC"],"primary_cat":"cs.LG","authors_text":"Liang Wang, Limin Xiao, Tianyang Jiang, Xiaojian Liao, Xiongwei Zhu, Yusen Zhang","submitted_at":"2026-05-26T14:32:56Z","abstract_excerpt":"Fine-grained Mixture-of-Experts (MoE) models sparsely activate only a subset of experts per token, reducing activated computation while maintaining high model capacity. However, in memory-constrained inference scenarios, only a small set of experts can be cached. Experts not in the cache must be fetched from slow external storage (e.g., UFS), leading to frequent evictions and substantial I/O overhead.\n  We propose ReMoE, a router fine-tuning framework designed to boost token-wise expert reuse. ReMoE biases the router toward recently selected experts, producing temporally stable routing that be"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.27081","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/2605.27081/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":"2605.27081","created_at":"2026-05-27T01:06:27.309013+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.27081v1","created_at":"2026-05-27T01:06:27.309013+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.27081","created_at":"2026-05-27T01:06:27.309013+00:00"},{"alias_kind":"pith_short_12","alias_value":"NE7W423ILRZW","created_at":"2026-05-27T01:06:27.309013+00:00"},{"alias_kind":"pith_short_16","alias_value":"NE7W423ILRZWFNV6","created_at":"2026-05-27T01:06:27.309013+00:00"},{"alias_kind":"pith_short_8","alias_value":"NE7W423I","created_at":"2026-05-27T01:06:27.309013+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/NE7W423ILRZWFNV6RN6FMRIGIU","json":"https://pith.science/pith/NE7W423ILRZWFNV6RN6FMRIGIU.json","graph_json":"https://pith.science/api/pith-number/NE7W423ILRZWFNV6RN6FMRIGIU/graph.json","events_json":"https://pith.science/api/pith-number/NE7W423ILRZWFNV6RN6FMRIGIU/events.json","paper":"https://pith.science/paper/NE7W423I"},"agent_actions":{"view_html":"https://pith.science/pith/NE7W423ILRZWFNV6RN6FMRIGIU","download_json":"https://pith.science/pith/NE7W423ILRZWFNV6RN6FMRIGIU.json","view_paper":"https://pith.science/paper/NE7W423I","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.27081&json=true","fetch_graph":"https://pith.science/api/pith-number/NE7W423ILRZWFNV6RN6FMRIGIU/graph.json","fetch_events":"https://pith.science/api/pith-number/NE7W423ILRZWFNV6RN6FMRIGIU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NE7W423ILRZWFNV6RN6FMRIGIU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NE7W423ILRZWFNV6RN6FMRIGIU/action/storage_attestation","attest_author":"https://pith.science/pith/NE7W423ILRZWFNV6RN6FMRIGIU/action/author_attestation","sign_citation":"https://pith.science/pith/NE7W423ILRZWFNV6RN6FMRIGIU/action/citation_signature","submit_replication":"https://pith.science/pith/NE7W423ILRZWFNV6RN6FMRIGIU/action/replication_record"}},"created_at":"2026-05-27T01:06:27.309013+00:00","updated_at":"2026-05-27T01:06:27.309013+00:00"}