{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:54ILOMC2WCXVKAB2A3C2E52PCE","short_pith_number":"pith:54ILOMC2","canonical_record":{"source":{"id":"2605.19049","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T19:14:52Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1e03518078c447808dfcbf153f4abdc5cd70d527271244f6d735e4fb8b167433","abstract_canon_sha256":"110238b62d1e097b59c7b0aaa0a6db96536584bcfbb6acec79d06093d72bdb4e"},"schema_version":"1.0"},"canonical_sha256":"ef10b7305ab0af55003a06c5a2774f1138f34100c0a40e30bdbf82ca3193011e","source":{"kind":"arxiv","id":"2605.19049","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.19049","created_at":"2026-05-20T01:05:24Z"},{"alias_kind":"arxiv_version","alias_value":"2605.19049v1","created_at":"2026-05-20T01:05:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.19049","created_at":"2026-05-20T01:05:24Z"},{"alias_kind":"pith_short_12","alias_value":"54ILOMC2WCXV","created_at":"2026-05-20T01:05:24Z"},{"alias_kind":"pith_short_16","alias_value":"54ILOMC2WCXVKAB2","created_at":"2026-05-20T01:05:24Z"},{"alias_kind":"pith_short_8","alias_value":"54ILOMC2","created_at":"2026-05-20T01:05:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:54ILOMC2WCXVKAB2A3C2E52PCE","target":"record","payload":{"canonical_record":{"source":{"id":"2605.19049","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T19:14:52Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1e03518078c447808dfcbf153f4abdc5cd70d527271244f6d735e4fb8b167433","abstract_canon_sha256":"110238b62d1e097b59c7b0aaa0a6db96536584bcfbb6acec79d06093d72bdb4e"},"schema_version":"1.0"},"canonical_sha256":"ef10b7305ab0af55003a06c5a2774f1138f34100c0a40e30bdbf82ca3193011e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:05:24.740044Z","signature_b64":"rGY8BuEmiPj0cz1SVvrCbUnfMUwXoPN7tTs706NPidhsHuYSWoyJWXsO4+k4rCIaajLniCRoIieBnLly757gDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ef10b7305ab0af55003a06c5a2774f1138f34100c0a40e30bdbf82ca3193011e","last_reissued_at":"2026-05-20T01:05:24.739502Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:05:24.739502Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.19049","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":"2026-05-20T01:05:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mJf18/kwhYgzkK5PN3xKFOtK/YdeCq0vCHNovS05ztye2acscERzz1SAIHkgVQm5dV1E7KShpi+N4Uy0TUUPAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T06:36:45.770013Z"},"content_sha256":"d3ab52b61e2dc366a352b5178b57bf55421440482776aea79e17bf2a9ad35001","schema_version":"1.0","event_id":"sha256:d3ab52b61e2dc366a352b5178b57bf55421440482776aea79e17bf2a9ad35001"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:54ILOMC2WCXVKAB2A3C2E52PCE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"KVBuffer: IO-aware Serving for Linear Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Lin Zhong, Longwei Zou","submitted_at":"2026-05-18T19:14:52Z","abstract_excerpt":"Linear attention has recently gained significant attention for long-context inference due to its constant decoding cost with respect to context length. However, existing serving systems typically serve linear attention by recurrently computing and updating a large linear attention state in every decoding step. Since the state is much larger than the per-token key and value, recurrent decoding incurs substantial memory access and becomes inefficient for serving linear attention. In this paper, we propose KVBuffer, an IO-aware serving mechanism for linear attention. By buffering recent keys and "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.19049","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/2605.19049/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":"2026-05-20T01:05:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LRjJALWcE/fCIUbkfPJ2DAfl/tM5oZNfMdXJx0TC7dSAHEVXcN44LUaHzeVju+V7xqae5cuFWqEMM/vHYGJyDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T06:36:45.770798Z"},"content_sha256":"d3a6ddb62b7a41e11ca02aed94f3a947f6f45a876656124fc6e2c76876f3e7d1","schema_version":"1.0","event_id":"sha256:d3a6ddb62b7a41e11ca02aed94f3a947f6f45a876656124fc6e2c76876f3e7d1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/54ILOMC2WCXVKAB2A3C2E52PCE/bundle.json","state_url":"https://pith.science/pith/54ILOMC2WCXVKAB2A3C2E52PCE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/54ILOMC2WCXVKAB2A3C2E52PCE/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":"2026-05-25T06:36:45Z","links":{"resolver":"https://pith.science/pith/54ILOMC2WCXVKAB2A3C2E52PCE","bundle":"https://pith.science/pith/54ILOMC2WCXVKAB2A3C2E52PCE/bundle.json","state":"https://pith.science/pith/54ILOMC2WCXVKAB2A3C2E52PCE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/54ILOMC2WCXVKAB2A3C2E52PCE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:54ILOMC2WCXVKAB2A3C2E52PCE","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":"110238b62d1e097b59c7b0aaa0a6db96536584bcfbb6acec79d06093d72bdb4e","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T19:14:52Z","title_canon_sha256":"1e03518078c447808dfcbf153f4abdc5cd70d527271244f6d735e4fb8b167433"},"schema_version":"1.0","source":{"id":"2605.19049","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.19049","created_at":"2026-05-20T01:05:24Z"},{"alias_kind":"arxiv_version","alias_value":"2605.19049v1","created_at":"2026-05-20T01:05:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.19049","created_at":"2026-05-20T01:05:24Z"},{"alias_kind":"pith_short_12","alias_value":"54ILOMC2WCXV","created_at":"2026-05-20T01:05:24Z"},{"alias_kind":"pith_short_16","alias_value":"54ILOMC2WCXVKAB2","created_at":"2026-05-20T01:05:24Z"},{"alias_kind":"pith_short_8","alias_value":"54ILOMC2","created_at":"2026-05-20T01:05:24Z"}],"graph_snapshots":[{"event_id":"sha256:d3a6ddb62b7a41e11ca02aed94f3a947f6f45a876656124fc6e2c76876f3e7d1","target":"graph","created_at":"2026-05-20T01:05:24Z","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/2605.19049/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Linear attention has recently gained significant attention for long-context inference due to its constant decoding cost with respect to context length. However, existing serving systems typically serve linear attention by recurrently computing and updating a large linear attention state in every decoding step. Since the state is much larger than the per-token key and value, recurrent decoding incurs substantial memory access and becomes inefficient for serving linear attention. In this paper, we propose KVBuffer, an IO-aware serving mechanism for linear attention. By buffering recent keys and ","authors_text":"Lin Zhong, Longwei Zou","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T19:14:52Z","title":"KVBuffer: IO-aware Serving for Linear Attention"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.19049","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:d3ab52b61e2dc366a352b5178b57bf55421440482776aea79e17bf2a9ad35001","target":"record","created_at":"2026-05-20T01:05:24Z","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":"110238b62d1e097b59c7b0aaa0a6db96536584bcfbb6acec79d06093d72bdb4e","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T19:14:52Z","title_canon_sha256":"1e03518078c447808dfcbf153f4abdc5cd70d527271244f6d735e4fb8b167433"},"schema_version":"1.0","source":{"id":"2605.19049","kind":"arxiv","version":1}},"canonical_sha256":"ef10b7305ab0af55003a06c5a2774f1138f34100c0a40e30bdbf82ca3193011e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ef10b7305ab0af55003a06c5a2774f1138f34100c0a40e30bdbf82ca3193011e","first_computed_at":"2026-05-20T01:05:24.739502Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T01:05:24.739502Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"rGY8BuEmiPj0cz1SVvrCbUnfMUwXoPN7tTs706NPidhsHuYSWoyJWXsO4+k4rCIaajLniCRoIieBnLly757gDg==","signature_status":"signed_v1","signed_at":"2026-05-20T01:05:24.740044Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.19049","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d3ab52b61e2dc366a352b5178b57bf55421440482776aea79e17bf2a9ad35001","sha256:d3a6ddb62b7a41e11ca02aed94f3a947f6f45a876656124fc6e2c76876f3e7d1"],"state_sha256":"2a8d384ae1a0fd7345723dfcec63e047cd9a26732010c198f6086906668d5b1a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YE0DphxDRGpxWX5Hi5lSHRJ0JBOUj3dqzM0L9JIJMfDJ24OYhndL86LN21N0Mxjs/blLHoObTFauYWRe92dvCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T06:36:45.774889Z","bundle_sha256":"31420d21bbabfbe99cc272269ab342678a3222e98275217f7ab3a8a5646e0a56"}}