{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:MSTAQJYSETO5T4UYL46I4JUXUZ","short_pith_number":"pith:MSTAQJYS","canonical_record":{"source":{"id":"2502.12085","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-17T17:59:56Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"7d2cc3291f76ebb8f8f56fc834cbfe87f7c75754b0067cb392a3adeed83f9bb5","abstract_canon_sha256":"65939a76e2d03bb59518f155021bd8c2a34148f56754fe4065fe5a9114b91959"},"schema_version":"1.0"},"canonical_sha256":"64a608271224ddd9f2985f3c8e2697a670346426bcec69bd3883fb201741a959","source":{"kind":"arxiv","id":"2502.12085","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.12085","created_at":"2026-07-05T11:09:45Z"},{"alias_kind":"arxiv_version","alias_value":"2502.12085v2","created_at":"2026-07-05T11:09:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.12085","created_at":"2026-07-05T11:09:45Z"},{"alias_kind":"pith_short_12","alias_value":"MSTAQJYSETO5","created_at":"2026-07-05T11:09:45Z"},{"alias_kind":"pith_short_16","alias_value":"MSTAQJYSETO5T4UY","created_at":"2026-07-05T11:09:45Z"},{"alias_kind":"pith_short_8","alias_value":"MSTAQJYS","created_at":"2026-07-05T11:09:45Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:MSTAQJYSETO5T4UYL46I4JUXUZ","target":"record","payload":{"canonical_record":{"source":{"id":"2502.12085","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-17T17:59:56Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"7d2cc3291f76ebb8f8f56fc834cbfe87f7c75754b0067cb392a3adeed83f9bb5","abstract_canon_sha256":"65939a76e2d03bb59518f155021bd8c2a34148f56754fe4065fe5a9114b91959"},"schema_version":"1.0"},"canonical_sha256":"64a608271224ddd9f2985f3c8e2697a670346426bcec69bd3883fb201741a959","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:09:45.362493Z","signature_b64":"DjbahK1UfrNClDTKrWfN6UgM7lEhkR8N2EGSt/UcTqzYIe3WAU0Vjy8uReFkpfctjsYYSe9CYRAHj4ja0QzjBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"64a608271224ddd9f2985f3c8e2697a670346426bcec69bd3883fb201741a959","last_reissued_at":"2026-07-05T11:09:45.361981Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:09:45.361981Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.12085","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:09:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"u+YC0T8ihoLMA3numxTpPRXlQLc3PDoMuKGjgzsgMXC4oxF6jA6vjctA5xbIa6XtcwPrrFVhNjfPw2wgH01OAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T22:49:14.931071Z"},"content_sha256":"5156b27333c99dcd86733dcdce4f17a3eb4f7fc947f2e5f6d97744e7b4bc47f1","schema_version":"1.0","event_id":"sha256:5156b27333c99dcd86733dcdce4f17a3eb4f7fc947f2e5f6d97744e7b4bc47f1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:MSTAQJYSETO5T4UYL46I4JUXUZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"APB: Accelerating Distributed Long-Context Inference by Passing Compressed Context Blocks across GPUs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Chaojun Xiao, Hao Zhou, Jie Zhou, Maosong Sun, Mingye Li, Sun Ao, Weilin Zhao, Xu Han, Yuxiang Huang, Zhiyuan Liu","submitted_at":"2025-02-17T17:59:56Z","abstract_excerpt":"While long-context inference is crucial for advancing large language model (LLM) applications, its prefill speed remains a significant bottleneck. Current approaches, including sequence parallelism strategies and compute reduction through approximate attention mechanisms, still fall short of delivering optimal inference efficiency. This hinders scaling the inputs to longer sequences and processing long-context queries in a timely manner. To address this, we introduce APB, an efficient long-context inference framework that leverages multi-host approximate attention to enhance prefill speed by r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.12085","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/2502.12085/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":"2026-07-05T11:09:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7Y9K7fgLELmX8ukwkk1/TXOrSXSaY9CGxAArY1WyYLJIeiLH6AdHkDWP1l5/fMS4wP3QJvNuJDyOegr+ba5SBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T22:49:14.931689Z"},"content_sha256":"fe78a3078ce2978f934a9b8fe25a71de23de641b949c6efb11d71fc300c590e7","schema_version":"1.0","event_id":"sha256:fe78a3078ce2978f934a9b8fe25a71de23de641b949c6efb11d71fc300c590e7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MSTAQJYSETO5T4UYL46I4JUXUZ/bundle.json","state_url":"https://pith.science/pith/MSTAQJYSETO5T4UYL46I4JUXUZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MSTAQJYSETO5T4UYL46I4JUXUZ/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":"2026-08-17T22:49:14Z","links":{"resolver":"https://pith.science/pith/MSTAQJYSETO5T4UYL46I4JUXUZ","bundle":"https://pith.science/pith/MSTAQJYSETO5T4UYL46I4JUXUZ/bundle.json","state":"https://pith.science/pith/MSTAQJYSETO5T4UYL46I4JUXUZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MSTAQJYSETO5T4UYL46I4JUXUZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:MSTAQJYSETO5T4UYL46I4JUXUZ","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":"65939a76e2d03bb59518f155021bd8c2a34148f56754fe4065fe5a9114b91959","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-17T17:59:56Z","title_canon_sha256":"7d2cc3291f76ebb8f8f56fc834cbfe87f7c75754b0067cb392a3adeed83f9bb5"},"schema_version":"1.0","source":{"id":"2502.12085","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.12085","created_at":"2026-07-05T11:09:45Z"},{"alias_kind":"arxiv_version","alias_value":"2502.12085v2","created_at":"2026-07-05T11:09:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.12085","created_at":"2026-07-05T11:09:45Z"},{"alias_kind":"pith_short_12","alias_value":"MSTAQJYSETO5","created_at":"2026-07-05T11:09:45Z"},{"alias_kind":"pith_short_16","alias_value":"MSTAQJYSETO5T4UY","created_at":"2026-07-05T11:09:45Z"},{"alias_kind":"pith_short_8","alias_value":"MSTAQJYS","created_at":"2026-07-05T11:09:45Z"}],"graph_snapshots":[{"event_id":"sha256:fe78a3078ce2978f934a9b8fe25a71de23de641b949c6efb11d71fc300c590e7","target":"graph","created_at":"2026-07-05T11:09:45Z","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/2502.12085/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"While long-context inference is crucial for advancing large language model (LLM) applications, its prefill speed remains a significant bottleneck. Current approaches, including sequence parallelism strategies and compute reduction through approximate attention mechanisms, still fall short of delivering optimal inference efficiency. This hinders scaling the inputs to longer sequences and processing long-context queries in a timely manner. To address this, we introduce APB, an efficient long-context inference framework that leverages multi-host approximate attention to enhance prefill speed by r","authors_text":"Chaojun Xiao, Hao Zhou, Jie Zhou, Maosong Sun, Mingye Li, Sun Ao, Weilin Zhao, Xu Han, Yuxiang Huang, Zhiyuan Liu","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-17T17:59:56Z","title":"APB: Accelerating Distributed Long-Context Inference by Passing Compressed Context Blocks across GPUs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.12085","kind":"arxiv","version":2},"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":"sha256:5156b27333c99dcd86733dcdce4f17a3eb4f7fc947f2e5f6d97744e7b4bc47f1","target":"record","created_at":"2026-07-05T11:09:45Z","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":"65939a76e2d03bb59518f155021bd8c2a34148f56754fe4065fe5a9114b91959","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-17T17:59:56Z","title_canon_sha256":"7d2cc3291f76ebb8f8f56fc834cbfe87f7c75754b0067cb392a3adeed83f9bb5"},"schema_version":"1.0","source":{"id":"2502.12085","kind":"arxiv","version":2}},"canonical_sha256":"64a608271224ddd9f2985f3c8e2697a670346426bcec69bd3883fb201741a959","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"64a608271224ddd9f2985f3c8e2697a670346426bcec69bd3883fb201741a959","first_computed_at":"2026-07-05T11:09:45.361981Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:09:45.361981Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"DjbahK1UfrNClDTKrWfN6UgM7lEhkR8N2EGSt/UcTqzYIe3WAU0Vjy8uReFkpfctjsYYSe9CYRAHj4ja0QzjBw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:09:45.362493Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.12085","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5156b27333c99dcd86733dcdce4f17a3eb4f7fc947f2e5f6d97744e7b4bc47f1","sha256:fe78a3078ce2978f934a9b8fe25a71de23de641b949c6efb11d71fc300c590e7"],"state_sha256":"cd0a1d2a3d36175651d01d8fb06aa76a9afb12952ed194472b70d13a41ec8353"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7jGf/6NilRim3d469tMmIlVaURYXdTJN9XQdVfgBIYs2Xk01v4moJondpgFQvOgpz7yhqV2ImAoOP3Dj6LZkBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-17T22:49:14.937339Z","bundle_sha256":"8f6aa9943b5e45494cb6d43c5e63bc0bdf649affab9d5229751127160c2cdb30"}}