{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:VFM4H2X4QLKUSESVGU6GVVGKAZ","short_pith_number":"pith:VFM4H2X4","schema_version":"1.0","canonical_sha256":"a959c3eafc82d5491255353c6ad4ca06570ca230a97c5fd18e1001760a710be7","source":{"kind":"arxiv","id":"2411.01433","version":2},"attestation_state":"computed","paper":{"title":"HOBBIT: A Mixed Precision Expert Offloading System for Fast MoE Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Chao Li, Jiacheng Liu, Jing Wang, Minyi Guo, Peng Tang, Pheng-Ann Heng, Xiaofeng Hou, Yifei Pu","submitted_at":"2024-11-03T04:25:46Z","abstract_excerpt":"The Mixture-of-Experts (MoE) architecture has demonstrated significant advantages in the era of Large Language Models (LLMs), offering enhanced capabilities with reduced inference costs. However, deploying MoE-based LLMs on memoryconstrained edge devices remains challenging due to their substantial memory requirements. While existing expertoffloading methods alleviate the memory requirements, they often incur significant expert-loading costs or compromise model accuracy. We present HOBBIT, a mixed precision expert offloading system to enable flexible and efficient MoE inference. Our key insigh"},"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":"2411.01433","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-11-03T04:25:46Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"6c349032c15c54eb777fee786eff34239ad046dffbaa612523cc846b2021df8f","abstract_canon_sha256":"bd55e1727e13f338c3188b7b4d0e6a63f99f32ba622dc38997c336164ce5e312"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:31:50.960551Z","signature_b64":"FYPeoKwnuhfb6Qjnv3NnqRj+covO2T+1JtbQrGNiUZlV7Gfhv/ts4kxekOest+97dpHz2jz/aMdGS/F8HUskDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a959c3eafc82d5491255353c6ad4ca06570ca230a97c5fd18e1001760a710be7","last_reissued_at":"2026-07-05T09:31:50.959999Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:31:50.959999Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HOBBIT: A Mixed Precision Expert Offloading System for Fast MoE Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Chao Li, Jiacheng Liu, Jing Wang, Minyi Guo, Peng Tang, Pheng-Ann Heng, Xiaofeng Hou, Yifei Pu","submitted_at":"2024-11-03T04:25:46Z","abstract_excerpt":"The Mixture-of-Experts (MoE) architecture has demonstrated significant advantages in the era of Large Language Models (LLMs), offering enhanced capabilities with reduced inference costs. However, deploying MoE-based LLMs on memoryconstrained edge devices remains challenging due to their substantial memory requirements. While existing expertoffloading methods alleviate the memory requirements, they often incur significant expert-loading costs or compromise model accuracy. We present HOBBIT, a mixed precision expert offloading system to enable flexible and efficient MoE inference. Our key insigh"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.01433","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/2411.01433/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":"2411.01433","created_at":"2026-07-05T09:31:50.960066+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.01433v2","created_at":"2026-07-05T09:31:50.960066+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.01433","created_at":"2026-07-05T09:31:50.960066+00:00"},{"alias_kind":"pith_short_12","alias_value":"VFM4H2X4QLKU","created_at":"2026-07-05T09:31:50.960066+00:00"},{"alias_kind":"pith_short_16","alias_value":"VFM4H2X4QLKUSESV","created_at":"2026-07-05T09:31:50.960066+00:00"},{"alias_kind":"pith_short_8","alias_value":"VFM4H2X4","created_at":"2026-07-05T09:31:50.960066+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.10493","citing_title":"Achieving Cloud-Grade SLOs for Local Mixture-of-Experts Inference through CPU-GPU Hybrid Design","ref_index":43,"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":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17889","citing_title":"CoX-MoE: Coalesced Expert Execution for High-Throughput MoE Inference with AMX-Enabled CPU-GPU Co-Execution","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2509.23638","citing_title":"LayerScope: Predictive Cross-Layer Scheduling for Efficient Multi-Batch MoE Inference on Legacy Servers","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11537","citing_title":"Fast MoE Inference via Predictive Prefetching and Expert Replication","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05819","citing_title":"HCInfer: An Efficient Inference System via Error Compensation for Resource-Constrained Devices","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05899","citing_title":"VisMMOE: Exploiting Visual-Expert Affinity for Efficient Visual-Language MoE Offloading","ref_index":46,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VFM4H2X4QLKUSESVGU6GVVGKAZ","json":"https://pith.science/pith/VFM4H2X4QLKUSESVGU6GVVGKAZ.json","graph_json":"https://pith.science/api/pith-number/VFM4H2X4QLKUSESVGU6GVVGKAZ/graph.json","events_json":"https://pith.science/api/pith-number/VFM4H2X4QLKUSESVGU6GVVGKAZ/events.json","paper":"https://pith.science/paper/VFM4H2X4"},"agent_actions":{"view_html":"https://pith.science/pith/VFM4H2X4QLKUSESVGU6GVVGKAZ","download_json":"https://pith.science/pith/VFM4H2X4QLKUSESVGU6GVVGKAZ.json","view_paper":"https://pith.science/paper/VFM4H2X4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.01433&json=true","fetch_graph":"https://pith.science/api/pith-number/VFM4H2X4QLKUSESVGU6GVVGKAZ/graph.json","fetch_events":"https://pith.science/api/pith-number/VFM4H2X4QLKUSESVGU6GVVGKAZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VFM4H2X4QLKUSESVGU6GVVGKAZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VFM4H2X4QLKUSESVGU6GVVGKAZ/action/storage_attestation","attest_author":"https://pith.science/pith/VFM4H2X4QLKUSESVGU6GVVGKAZ/action/author_attestation","sign_citation":"https://pith.science/pith/VFM4H2X4QLKUSESVGU6GVVGKAZ/action/citation_signature","submit_replication":"https://pith.science/pith/VFM4H2X4QLKUSESVGU6GVVGKAZ/action/replication_record"}},"created_at":"2026-07-05T09:31:50.960066+00:00","updated_at":"2026-07-05T09:31:50.960066+00:00"}