{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:TEBNBAJX7PEIHNUD72JKUYTYLD","short_pith_number":"pith:TEBNBAJX","schema_version":"1.0","canonical_sha256":"9902d08137fbc883b683fe92aa627858f7a88ff5c74274b99e81ec36071e27e7","source":{"kind":"arxiv","id":"2606.29207","version":1},"attestation_state":"computed","paper":{"title":"KernelFlume: Elastic Core-Attention Scaling for Agentic Long-Context Decoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Guangyu Xiang, Lin Zhang, Shaohuai Shi, Wenxiang Lin, Xiaowen Chu, Xueze Kang, Yuxin Wang","submitted_at":"2026-06-28T05:16:17Z","abstract_excerpt":"LLM serving is increasingly dominated by long and dynamic decode workloads from agents, reasoning models, and extended conversations. When bursty long-context demand exceeds deployed capacity, existing serving systems typically scale out by launching additional serving instances with model replicas. This instance-level elasticity increases KV capacity only by provisioning another full copy of the model, inheriting startup latency, memory overhead, and batch fragmentation.\n  We present KernelFlume, a decode-centric architecture that disaggregates the stable projection/FFN path from core-attenti"},"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":"2606.29207","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-06-28T05:16:17Z","cross_cats_sorted":[],"title_canon_sha256":"31701c0f5e1897f7351c894f87e3f7eb42b854d17a066393d5290ab68fe1d708","abstract_canon_sha256":"f834b39279a8655799ec2e0a33a7ba301defc74a6ea6b5c0257a3bb33d7af1c0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T01:17:57.227901Z","signature_b64":"Fd+PHAVnFKBb8YgdzlaZqODf2X7wCv7GzioeHD0boZ7CwErSjPXv+defMVleK7isBX2luXjjbD0cwFI6ZaXqCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9902d08137fbc883b683fe92aa627858f7a88ff5c74274b99e81ec36071e27e7","last_reissued_at":"2026-06-30T01:17:57.227423Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T01:17:57.227423Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"KernelFlume: Elastic Core-Attention Scaling for Agentic Long-Context Decoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Guangyu Xiang, Lin Zhang, Shaohuai Shi, Wenxiang Lin, Xiaowen Chu, Xueze Kang, Yuxin Wang","submitted_at":"2026-06-28T05:16:17Z","abstract_excerpt":"LLM serving is increasingly dominated by long and dynamic decode workloads from agents, reasoning models, and extended conversations. When bursty long-context demand exceeds deployed capacity, existing serving systems typically scale out by launching additional serving instances with model replicas. This instance-level elasticity increases KV capacity only by provisioning another full copy of the model, inheriting startup latency, memory overhead, and batch fragmentation.\n  We present KernelFlume, a decode-centric architecture that disaggregates the stable projection/FFN path from core-attenti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.29207","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/2606.29207/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":"2606.29207","created_at":"2026-06-30T01:17:57.227490+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.29207v1","created_at":"2026-06-30T01:17:57.227490+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.29207","created_at":"2026-06-30T01:17:57.227490+00:00"},{"alias_kind":"pith_short_12","alias_value":"TEBNBAJX7PEI","created_at":"2026-06-30T01:17:57.227490+00:00"},{"alias_kind":"pith_short_16","alias_value":"TEBNBAJX7PEIHNUD","created_at":"2026-06-30T01:17:57.227490+00:00"},{"alias_kind":"pith_short_8","alias_value":"TEBNBAJX","created_at":"2026-06-30T01:17:57.227490+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/TEBNBAJX7PEIHNUD72JKUYTYLD","json":"https://pith.science/pith/TEBNBAJX7PEIHNUD72JKUYTYLD.json","graph_json":"https://pith.science/api/pith-number/TEBNBAJX7PEIHNUD72JKUYTYLD/graph.json","events_json":"https://pith.science/api/pith-number/TEBNBAJX7PEIHNUD72JKUYTYLD/events.json","paper":"https://pith.science/paper/TEBNBAJX"},"agent_actions":{"view_html":"https://pith.science/pith/TEBNBAJX7PEIHNUD72JKUYTYLD","download_json":"https://pith.science/pith/TEBNBAJX7PEIHNUD72JKUYTYLD.json","view_paper":"https://pith.science/paper/TEBNBAJX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.29207&json=true","fetch_graph":"https://pith.science/api/pith-number/TEBNBAJX7PEIHNUD72JKUYTYLD/graph.json","fetch_events":"https://pith.science/api/pith-number/TEBNBAJX7PEIHNUD72JKUYTYLD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TEBNBAJX7PEIHNUD72JKUYTYLD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TEBNBAJX7PEIHNUD72JKUYTYLD/action/storage_attestation","attest_author":"https://pith.science/pith/TEBNBAJX7PEIHNUD72JKUYTYLD/action/author_attestation","sign_citation":"https://pith.science/pith/TEBNBAJX7PEIHNUD72JKUYTYLD/action/citation_signature","submit_replication":"https://pith.science/pith/TEBNBAJX7PEIHNUD72JKUYTYLD/action/replication_record"}},"created_at":"2026-06-30T01:17:57.227490+00:00","updated_at":"2026-06-30T01:17:57.227490+00:00"}