{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:AGLF3HR23W3Q6OTVZW56EAPVRD","short_pith_number":"pith:AGLF3HR2","canonical_record":{"source":{"id":"2507.07120","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2025-07-07T19:47:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8b271adb311358672273c1049c38793c5d0171a83beb61e70a59050d32d31bb8","abstract_canon_sha256":"d1eb31b7816ca374c8c99ab8c7550ec2eef974dd104299763ce21a95db11f72f"},"schema_version":"1.0"},"canonical_sha256":"01965d9e3addb70f3a75cdbbe201f588ff33434768e71ca9743b9d26ff2a1308","source":{"kind":"arxiv","id":"2507.07120","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.07120","created_at":"2026-07-05T11:34:31Z"},{"alias_kind":"arxiv_version","alias_value":"2507.07120v1","created_at":"2026-07-05T11:34:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.07120","created_at":"2026-07-05T11:34:31Z"},{"alias_kind":"pith_short_12","alias_value":"AGLF3HR23W3Q","created_at":"2026-07-05T11:34:31Z"},{"alias_kind":"pith_short_16","alias_value":"AGLF3HR23W3Q6OTV","created_at":"2026-07-05T11:34:31Z"},{"alias_kind":"pith_short_8","alias_value":"AGLF3HR2","created_at":"2026-07-05T11:34:31Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:AGLF3HR23W3Q6OTVZW56EAPVRD","target":"record","payload":{"canonical_record":{"source":{"id":"2507.07120","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2025-07-07T19:47:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8b271adb311358672273c1049c38793c5d0171a83beb61e70a59050d32d31bb8","abstract_canon_sha256":"d1eb31b7816ca374c8c99ab8c7550ec2eef974dd104299763ce21a95db11f72f"},"schema_version":"1.0"},"canonical_sha256":"01965d9e3addb70f3a75cdbbe201f588ff33434768e71ca9743b9d26ff2a1308","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:34:31.302325Z","signature_b64":"g2f8B5fFS60zgB+lhMtE3s/rSS6e44sz49xuvYddjj2MM5vLrILnsXm3HVq4pobhXYNqF0GsYh7LRLgBtehDDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"01965d9e3addb70f3a75cdbbe201f588ff33434768e71ca9743b9d26ff2a1308","last_reissued_at":"2026-07-05T11:34:31.301851Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:34:31.301851Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2507.07120","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":"2026-07-05T11:34:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HxVLis2WObQWRsHLTLyfCKqGspcyy+p6bgc4d3DPJVGUoxw9TJwpy9y3J6WPgIcNjDk0QE9RUpIYDM1T3MbhDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T15:25:50.970719Z"},"content_sha256":"bdca602ce11ae365e3edd6739328da518c8bd8e671522a6533f9eb2b70b07eae","schema_version":"1.0","event_id":"sha256:bdca602ce11ae365e3edd6739328da518c8bd8e671522a6533f9eb2b70b07eae"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:AGLF3HR23W3Q6OTVZW56EAPVRD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Helix Parallelism: Rethinking Sharding Strategies for Interactive Multi-Million-Token LLM Decoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.DC","authors_text":"Ankit More, Bita Darvish Rouhani, Brian Pharris, Dheevatsa Mudigere, Maximilian Golub, Nidhi Bhatia, Ramon Matas, Ritchie Zhao, Ritika Borkar, Tiyasa Mitra","submitted_at":"2025-07-07T19:47:24Z","abstract_excerpt":"As LLMs scale to multi-million-token KV histories, real-time autoregressive decoding under tight Token-to-Token Latency (TTL) constraints faces growing pressure. Two core bottlenecks dominate: accessing Feed-Forward Network (FFN) weights and reading long KV caches. While Tensor Parallelism (TP) helps mitigate the cost of FFN weight reads, it does not scale well for attention. When TP width exceeds the number of KV heads, it leads to inefficient KV duplication, limits parallelism, and constrains batch size. Simultaneously, DRAM reads for long KV histories scale linearly with batch size, further"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.07120","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/2507.07120/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":"2026-07-05T11:34:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9xa5o48jmBjSu/UQDbKrwsdHHMZPxoeMtjgPXHkWthy/nrEqodOImcy1b9uplW5SqE8OBfF9FYXdvPMXtjwOAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T15:25:50.971287Z"},"content_sha256":"6e83ae9790a56376074855d993202b862a302470b1014abdb5abc47967f86f46","schema_version":"1.0","event_id":"sha256:6e83ae9790a56376074855d993202b862a302470b1014abdb5abc47967f86f46"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/AGLF3HR23W3Q6OTVZW56EAPVRD/bundle.json","state_url":"https://pith.science/pith/AGLF3HR23W3Q6OTVZW56EAPVRD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/AGLF3HR23W3Q6OTVZW56EAPVRD/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":"2026-08-20T15:25:50Z","links":{"resolver":"https://pith.science/pith/AGLF3HR23W3Q6OTVZW56EAPVRD","bundle":"https://pith.science/pith/AGLF3HR23W3Q6OTVZW56EAPVRD/bundle.json","state":"https://pith.science/pith/AGLF3HR23W3Q6OTVZW56EAPVRD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/AGLF3HR23W3Q6OTVZW56EAPVRD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:AGLF3HR23W3Q6OTVZW56EAPVRD","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":"d1eb31b7816ca374c8c99ab8c7550ec2eef974dd104299763ce21a95db11f72f","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2025-07-07T19:47:24Z","title_canon_sha256":"8b271adb311358672273c1049c38793c5d0171a83beb61e70a59050d32d31bb8"},"schema_version":"1.0","source":{"id":"2507.07120","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.07120","created_at":"2026-07-05T11:34:31Z"},{"alias_kind":"arxiv_version","alias_value":"2507.07120v1","created_at":"2026-07-05T11:34:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.07120","created_at":"2026-07-05T11:34:31Z"},{"alias_kind":"pith_short_12","alias_value":"AGLF3HR23W3Q","created_at":"2026-07-05T11:34:31Z"},{"alias_kind":"pith_short_16","alias_value":"AGLF3HR23W3Q6OTV","created_at":"2026-07-05T11:34:31Z"},{"alias_kind":"pith_short_8","alias_value":"AGLF3HR2","created_at":"2026-07-05T11:34:31Z"}],"graph_snapshots":[{"event_id":"sha256:6e83ae9790a56376074855d993202b862a302470b1014abdb5abc47967f86f46","target":"graph","created_at":"2026-07-05T11:34:31Z","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/2507.07120/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As LLMs scale to multi-million-token KV histories, real-time autoregressive decoding under tight Token-to-Token Latency (TTL) constraints faces growing pressure. Two core bottlenecks dominate: accessing Feed-Forward Network (FFN) weights and reading long KV caches. While Tensor Parallelism (TP) helps mitigate the cost of FFN weight reads, it does not scale well for attention. When TP width exceeds the number of KV heads, it leads to inefficient KV duplication, limits parallelism, and constrains batch size. Simultaneously, DRAM reads for long KV histories scale linearly with batch size, further","authors_text":"Ankit More, Bita Darvish Rouhani, Brian Pharris, Dheevatsa Mudigere, Maximilian Golub, Nidhi Bhatia, Ramon Matas, Ritchie Zhao, Ritika Borkar, Tiyasa Mitra","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2025-07-07T19:47:24Z","title":"Helix Parallelism: Rethinking Sharding Strategies for Interactive Multi-Million-Token LLM Decoding"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.07120","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:bdca602ce11ae365e3edd6739328da518c8bd8e671522a6533f9eb2b70b07eae","target":"record","created_at":"2026-07-05T11:34:31Z","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":"d1eb31b7816ca374c8c99ab8c7550ec2eef974dd104299763ce21a95db11f72f","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2025-07-07T19:47:24Z","title_canon_sha256":"8b271adb311358672273c1049c38793c5d0171a83beb61e70a59050d32d31bb8"},"schema_version":"1.0","source":{"id":"2507.07120","kind":"arxiv","version":1}},"canonical_sha256":"01965d9e3addb70f3a75cdbbe201f588ff33434768e71ca9743b9d26ff2a1308","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"01965d9e3addb70f3a75cdbbe201f588ff33434768e71ca9743b9d26ff2a1308","first_computed_at":"2026-07-05T11:34:31.301851Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:34:31.301851Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"g2f8B5fFS60zgB+lhMtE3s/rSS6e44sz49xuvYddjj2MM5vLrILnsXm3HVq4pobhXYNqF0GsYh7LRLgBtehDDw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:34:31.302325Z","signed_message":"canonical_sha256_bytes"},"source_id":"2507.07120","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bdca602ce11ae365e3edd6739328da518c8bd8e671522a6533f9eb2b70b07eae","sha256:6e83ae9790a56376074855d993202b862a302470b1014abdb5abc47967f86f46"],"state_sha256":"3564908cd350166d73fd7534bad80d8128510ea520f907878fa248219a4247eb"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"AddE8a3LHsHLcOMUxRQ3bzZlXejZWfYIEhBTV2UKAY7ttWYxknhsg2n2ysnC+vNY7+SiVfFuJjabH+FJSDL4Bg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-20T15:25:50.976716Z","bundle_sha256":"9893622008dd7c3c5b3b712833c5e8a33572e40257d8b7f707eee87e3cb914b6"}}