{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:EL6A4QOTPLYS3LFEIZGOL5QVTM","short_pith_number":"pith:EL6A4QOT","canonical_record":{"source":{"id":"2606.26666","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T06:56:43Z","cross_cats_sorted":[],"title_canon_sha256":"9d8c38979e0d02daad360388d7246e3f51d9bd088c2a886a7ffc1d58b95f9116","abstract_canon_sha256":"6586281ebd085d9f0661bdf4e48c8899bb25db46360150cc184174c62866661d"},"schema_version":"1.0"},"canonical_sha256":"22fc0e41d37af12daca4464ce5f6159b3b7926b56310533ae0f6d3b869338159","source":{"kind":"arxiv","id":"2606.26666","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.26666","created_at":"2026-06-26T01:15:56Z"},{"alias_kind":"arxiv_version","alias_value":"2606.26666v1","created_at":"2026-06-26T01:15:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.26666","created_at":"2026-06-26T01:15:56Z"},{"alias_kind":"pith_short_12","alias_value":"EL6A4QOTPLYS","created_at":"2026-06-26T01:15:56Z"},{"alias_kind":"pith_short_16","alias_value":"EL6A4QOTPLYS3LFE","created_at":"2026-06-26T01:15:56Z"},{"alias_kind":"pith_short_8","alias_value":"EL6A4QOT","created_at":"2026-06-26T01:15:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:EL6A4QOTPLYS3LFEIZGOL5QVTM","target":"record","payload":{"canonical_record":{"source":{"id":"2606.26666","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T06:56:43Z","cross_cats_sorted":[],"title_canon_sha256":"9d8c38979e0d02daad360388d7246e3f51d9bd088c2a886a7ffc1d58b95f9116","abstract_canon_sha256":"6586281ebd085d9f0661bdf4e48c8899bb25db46360150cc184174c62866661d"},"schema_version":"1.0"},"canonical_sha256":"22fc0e41d37af12daca4464ce5f6159b3b7926b56310533ae0f6d3b869338159","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-26T01:15:56.296492Z","signature_b64":"iURSKD0bjYIRaZXYb6MLzIBS7G/8Gukz0Vl5XfHevf6Db05tBVD3j39rv6Sf7MNoDdbz5KOIBHI7F7UW13npBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"22fc0e41d37af12daca4464ce5f6159b3b7926b56310533ae0f6d3b869338159","last_reissued_at":"2026-06-26T01:15:56.296077Z","signature_status":"signed_v1","first_computed_at":"2026-06-26T01:15:56.296077Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.26666","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-26T01:15:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+JEwXVh9oPmo8coszJ6GrkjZI+14MFV1trsB3VwO+HXFcWxlDr0n/BN04cBAq37kjwcnLaYjlHNfeqlpvrX1Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-03T09:59:00.310931Z"},"content_sha256":"0114ee578c80f1450e5ceb2546ece1914c4d78ac36b32d12544b84d921dc9dbc","schema_version":"1.0","event_id":"sha256:0114ee578c80f1450e5ceb2546ece1914c4d78ac36b32d12544b84d921dc9dbc"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:EL6A4QOTPLYS3LFEIZGOL5QVTM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"PersistentKV: Page-Aware Decode Scheduling for Long-Context LLM Serving on Commodity GPUs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Muhammad Ahmed","submitted_at":"2026-06-25T06:56:43Z","abstract_excerpt":"Autoregressive large language model (LLM) serving is increasingly limited by key-value (KV) cache movement rather than dense matrix multiplication. Modern paged-attention systems reduce KV-cache fragmentation and mature kernels such as FlashInfer provide highly optimized native-paged decode attention. However, the best single-kernel implementation is not always the best serving schedule: low-active long-context decode can under-utilize commodity GPUs, while mixed sequence lengths introduce a tension between many exact-length launches and coarse padded batches. We present PersistentKV, a native"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.26666","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/2606.26666/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-26T01:15:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BwkX1uCYCyry9+VQriBo19Us62uO4IEPjrkfuHBpUxZ2FwvawfLQmbZRExRRGtzf2BHgy9K5r0BqwmGoN15VCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-03T09:59:00.311301Z"},"content_sha256":"3a12c85cac8bbe787202fc8e2f71c299a14e6d0b4ee10396285488d2879e7eb0","schema_version":"1.0","event_id":"sha256:3a12c85cac8bbe787202fc8e2f71c299a14e6d0b4ee10396285488d2879e7eb0"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/EL6A4QOTPLYS3LFEIZGOL5QVTM/bundle.json","state_url":"https://pith.science/pith/EL6A4QOTPLYS3LFEIZGOL5QVTM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/EL6A4QOTPLYS3LFEIZGOL5QVTM/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-07-03T09:59:00Z","links":{"resolver":"https://pith.science/pith/EL6A4QOTPLYS3LFEIZGOL5QVTM","bundle":"https://pith.science/pith/EL6A4QOTPLYS3LFEIZGOL5QVTM/bundle.json","state":"https://pith.science/pith/EL6A4QOTPLYS3LFEIZGOL5QVTM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/EL6A4QOTPLYS3LFEIZGOL5QVTM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:EL6A4QOTPLYS3LFEIZGOL5QVTM","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":"6586281ebd085d9f0661bdf4e48c8899bb25db46360150cc184174c62866661d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T06:56:43Z","title_canon_sha256":"9d8c38979e0d02daad360388d7246e3f51d9bd088c2a886a7ffc1d58b95f9116"},"schema_version":"1.0","source":{"id":"2606.26666","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.26666","created_at":"2026-06-26T01:15:56Z"},{"alias_kind":"arxiv_version","alias_value":"2606.26666v1","created_at":"2026-06-26T01:15:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.26666","created_at":"2026-06-26T01:15:56Z"},{"alias_kind":"pith_short_12","alias_value":"EL6A4QOTPLYS","created_at":"2026-06-26T01:15:56Z"},{"alias_kind":"pith_short_16","alias_value":"EL6A4QOTPLYS3LFE","created_at":"2026-06-26T01:15:56Z"},{"alias_kind":"pith_short_8","alias_value":"EL6A4QOT","created_at":"2026-06-26T01:15:56Z"}],"graph_snapshots":[{"event_id":"sha256:3a12c85cac8bbe787202fc8e2f71c299a14e6d0b4ee10396285488d2879e7eb0","target":"graph","created_at":"2026-06-26T01:15:56Z","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/2606.26666/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Autoregressive large language model (LLM) serving is increasingly limited by key-value (KV) cache movement rather than dense matrix multiplication. Modern paged-attention systems reduce KV-cache fragmentation and mature kernels such as FlashInfer provide highly optimized native-paged decode attention. However, the best single-kernel implementation is not always the best serving schedule: low-active long-context decode can under-utilize commodity GPUs, while mixed sequence lengths introduce a tension between many exact-length launches and coarse padded batches. We present PersistentKV, a native","authors_text":"Muhammad Ahmed","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T06:56:43Z","title":"PersistentKV: Page-Aware Decode Scheduling for Long-Context LLM Serving on Commodity GPUs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.26666","kind":"arxiv","version":1},"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":"sha256:0114ee578c80f1450e5ceb2546ece1914c4d78ac36b32d12544b84d921dc9dbc","target":"record","created_at":"2026-06-26T01:15:56Z","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":"6586281ebd085d9f0661bdf4e48c8899bb25db46360150cc184174c62866661d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T06:56:43Z","title_canon_sha256":"9d8c38979e0d02daad360388d7246e3f51d9bd088c2a886a7ffc1d58b95f9116"},"schema_version":"1.0","source":{"id":"2606.26666","kind":"arxiv","version":1}},"canonical_sha256":"22fc0e41d37af12daca4464ce5f6159b3b7926b56310533ae0f6d3b869338159","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"22fc0e41d37af12daca4464ce5f6159b3b7926b56310533ae0f6d3b869338159","first_computed_at":"2026-06-26T01:15:56.296077Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-26T01:15:56.296077Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"iURSKD0bjYIRaZXYb6MLzIBS7G/8Gukz0Vl5XfHevf6Db05tBVD3j39rv6Sf7MNoDdbz5KOIBHI7F7UW13npBA==","signature_status":"signed_v1","signed_at":"2026-06-26T01:15:56.296492Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.26666","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0114ee578c80f1450e5ceb2546ece1914c4d78ac36b32d12544b84d921dc9dbc","sha256:3a12c85cac8bbe787202fc8e2f71c299a14e6d0b4ee10396285488d2879e7eb0"],"state_sha256":"cfe1261c0ec9563426c531fa715bd7f53ded5921c755420c3e91715f93542aa6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rzDpq1O+GdVIaf/oxd/a7xAISv6EVIbw0X5h823bR+KH2TFZ7WHw9P+I9k+YFdaRZ1rEH/LhYmfq6/+CZ469Cw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-03T09:59:00.313285Z","bundle_sha256":"3b5f0d0ce8e785b95268a03cd5cb1f802dcce66b024d60622c2d5e42ada6c77e"}}