{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:GTP3RAT2XIBXGXMA3THUE7IO23","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"ee86c1298bc4029c886d846053e078160b531ee6b29728429348026c1cde3983","cross_cats_sorted":["cs.PF"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-01-25T18:07:50Z","title_canon_sha256":"d6e5f0c52d751d32acd500f3de11e0978d767c13f6460f7bc95107d83df9fb1a"},"schema_version":"1.0","source":{"id":"2401.14361","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.14361","created_at":"2026-07-05T10:30:00Z"},{"alias_kind":"arxiv_version","alias_value":"2401.14361v3","created_at":"2026-07-05T10:30:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.14361","created_at":"2026-07-05T10:30:00Z"},{"alias_kind":"pith_short_12","alias_value":"GTP3RAT2XIBX","created_at":"2026-07-05T10:30:00Z"},{"alias_kind":"pith_short_16","alias_value":"GTP3RAT2XIBXGXMA","created_at":"2026-07-05T10:30:00Z"},{"alias_kind":"pith_short_8","alias_value":"GTP3RAT2","created_at":"2026-07-05T10:30:00Z"}],"graph_snapshots":[{"event_id":"sha256:56ffc7e6fba6ef3c92278f7b0c58a142098f888ad2831168fb5f187ca83ebbc9","target":"graph","created_at":"2026-07-05T10:30:00Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2401.14361/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper presents MoE-Infinity, an efficient MoE inference system designed for personal machines with limited GPU memory capacity. The key idea for MoE-Infinity is that on personal machines, which are often single-user environments, MoE-based LLMs typically operate with a batch size of one. In this setting, MoE models exhibit a high degree of activation sparsity, meaning a small number of experts are frequently reused in generating tokens during the decode phase. Leveraging this idea, we design a sparsity-aware expert cache, which can trace the sparse activation of experts during inference a","authors_text":"Leyang Xue, Luo Mai, Mahesh Marina, Yao Fu, Zhan Lu","cross_cats":["cs.PF"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-01-25T18:07:50Z","title":"MoE-Infinity: Efficient MoE Inference on Personal Machines with Sparsity-Aware Expert Cache"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.14361","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:c8e99747a681f3518ec8071df875eb1c27ef7b37751e73182c2a50755eb534b6","target":"record","created_at":"2026-07-05T10:30:00Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"ee86c1298bc4029c886d846053e078160b531ee6b29728429348026c1cde3983","cross_cats_sorted":["cs.PF"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-01-25T18:07:50Z","title_canon_sha256":"d6e5f0c52d751d32acd500f3de11e0978d767c13f6460f7bc95107d83df9fb1a"},"schema_version":"1.0","source":{"id":"2401.14361","kind":"arxiv","version":3}},"canonical_sha256":"34dfb8827aba03735d80dccf427d0ed6d200f37a370322e153918b8fc8729cee","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"34dfb8827aba03735d80dccf427d0ed6d200f37a370322e153918b8fc8729cee","first_computed_at":"2026-07-05T10:30:00.452818Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:30:00.452818Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"eCRNCQSC0CjP29PUrvUDIXHSzwuWuuN0FL7QUipICrU2ySmgNlV7ecVPyaedIt/+/BltmKwIBB9rEZgGFvGxBA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:30:00.453363Z","signed_message":"canonical_sha256_bytes"},"source_id":"2401.14361","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c8e99747a681f3518ec8071df875eb1c27ef7b37751e73182c2a50755eb534b6","sha256:56ffc7e6fba6ef3c92278f7b0c58a142098f888ad2831168fb5f187ca83ebbc9"],"state_sha256":"6ee32fc1ab4db508d32aff0f0c2d2a9f8bb68cd3574f053ef563f6ff48383e0f"}