{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:2N5EE55GNF5TG43HSNPDXID2S2","short_pith_number":"pith:2N5EE55G","canonical_record":{"source":{"id":"2505.01099","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-02T08:23:29Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"678cb8153bcb9c5a52d3c4f59365687d5f0684a2113729f7a78ca3b8e3ac86b7","abstract_canon_sha256":"189ddfab2afc990fc9336a42e3fab6ecadc28d6fa22ba310864914d12586f6cc"},"schema_version":"1.0"},"canonical_sha256":"d37a4277a6697b337367935e3ba07a96b780a50c64df857010b27309570af48e","source":{"kind":"arxiv","id":"2505.01099","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.01099","created_at":"2026-07-05T10:57:45Z"},{"alias_kind":"arxiv_version","alias_value":"2505.01099v1","created_at":"2026-07-05T10:57:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.01099","created_at":"2026-07-05T10:57:45Z"},{"alias_kind":"pith_short_12","alias_value":"2N5EE55GNF5T","created_at":"2026-07-05T10:57:45Z"},{"alias_kind":"pith_short_16","alias_value":"2N5EE55GNF5TG43H","created_at":"2026-07-05T10:57:45Z"},{"alias_kind":"pith_short_8","alias_value":"2N5EE55G","created_at":"2026-07-05T10:57:45Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:2N5EE55GNF5TG43HSNPDXID2S2","target":"record","payload":{"canonical_record":{"source":{"id":"2505.01099","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-02T08:23:29Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"678cb8153bcb9c5a52d3c4f59365687d5f0684a2113729f7a78ca3b8e3ac86b7","abstract_canon_sha256":"189ddfab2afc990fc9336a42e3fab6ecadc28d6fa22ba310864914d12586f6cc"},"schema_version":"1.0"},"canonical_sha256":"d37a4277a6697b337367935e3ba07a96b780a50c64df857010b27309570af48e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:57:45.107934Z","signature_b64":"vl0GMiv7nmwbmuAxodL1A/mruUdIOwxdum2lkdAgnPVQX3HqaYNAIYV96ZE6S/aWJufncoBsSIgXSu2fmC6zBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d37a4277a6697b337367935e3ba07a96b780a50c64df857010b27309570af48e","last_reissued_at":"2026-07-05T10:57:45.107404Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:57:45.107404Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.01099","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":"2026-07-05T10:57:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CxL83YcGi2u2D/L+iceZJTk3gaekRNDMhxEXtBM8/DhBu+Sm/5By5jfrX+CsYV8+IEvDfksqiyZWSWooFinEBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T06:44:27.166366Z"},"content_sha256":"06b27c4d89b0bbba6d2656c5b71150619105dc0b264eaae973c036dd11df942e","schema_version":"1.0","event_id":"sha256:06b27c4d89b0bbba6d2656c5b71150619105dc0b264eaae973c036dd11df942e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:2N5EE55GNF5TG43HSNPDXID2S2","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Nesterov Method for Asynchronous Pipeline Parallel Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Alexander Long, Gil Avraham, Sameera Ramasinghe, Thalaiyasingam Ajanthan, Yan Zuo","submitted_at":"2025-05-02T08:23:29Z","abstract_excerpt":"Pipeline Parallelism (PP) enables large neural network training on small, interconnected devices by splitting the model into multiple stages. To maximize pipeline utilization, asynchronous optimization is appealing as it offers 100% pipeline utilization by construction. However, it is inherently challenging as the weights and gradients are no longer synchronized, leading to stale (or delayed) gradients. To alleviate this, we introduce a variant of Nesterov Accelerated Gradient (NAG) for asynchronous optimization in PP. Specifically, we modify the look-ahead step in NAG to effectively address t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.01099","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/2505.01099/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":"2026-07-05T10:57:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VkHf/AmJ4E1ix9QtOTlkPiVNfg89odW6lk+bjVHgah8y5GXanM4QL+lcPfCGf1m1b7pMZ4Jj82X/0UZBbakaBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T06:44:27.166909Z"},"content_sha256":"a10ee05072e6ff21c758b676b85f24e0d2f7e39cf5cf8a78de87568c5a7e8875","schema_version":"1.0","event_id":"sha256:a10ee05072e6ff21c758b676b85f24e0d2f7e39cf5cf8a78de87568c5a7e8875"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/2N5EE55GNF5TG43HSNPDXID2S2/bundle.json","state_url":"https://pith.science/pith/2N5EE55GNF5TG43HSNPDXID2S2/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/2N5EE55GNF5TG43HSNPDXID2S2/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":"2026-08-18T06:44:27Z","links":{"resolver":"https://pith.science/pith/2N5EE55GNF5TG43HSNPDXID2S2","bundle":"https://pith.science/pith/2N5EE55GNF5TG43HSNPDXID2S2/bundle.json","state":"https://pith.science/pith/2N5EE55GNF5TG43HSNPDXID2S2/state.json","well_known_bundle":"https://pith.science/.well-known/pith/2N5EE55GNF5TG43HSNPDXID2S2/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:2N5EE55GNF5TG43HSNPDXID2S2","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":"189ddfab2afc990fc9336a42e3fab6ecadc28d6fa22ba310864914d12586f6cc","cross_cats_sorted":["cs.DC"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-02T08:23:29Z","title_canon_sha256":"678cb8153bcb9c5a52d3c4f59365687d5f0684a2113729f7a78ca3b8e3ac86b7"},"schema_version":"1.0","source":{"id":"2505.01099","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.01099","created_at":"2026-07-05T10:57:45Z"},{"alias_kind":"arxiv_version","alias_value":"2505.01099v1","created_at":"2026-07-05T10:57:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.01099","created_at":"2026-07-05T10:57:45Z"},{"alias_kind":"pith_short_12","alias_value":"2N5EE55GNF5T","created_at":"2026-07-05T10:57:45Z"},{"alias_kind":"pith_short_16","alias_value":"2N5EE55GNF5TG43H","created_at":"2026-07-05T10:57:45Z"},{"alias_kind":"pith_short_8","alias_value":"2N5EE55G","created_at":"2026-07-05T10:57:45Z"}],"graph_snapshots":[{"event_id":"sha256:a10ee05072e6ff21c758b676b85f24e0d2f7e39cf5cf8a78de87568c5a7e8875","target":"graph","created_at":"2026-07-05T10:57:45Z","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/2505.01099/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Pipeline Parallelism (PP) enables large neural network training on small, interconnected devices by splitting the model into multiple stages. To maximize pipeline utilization, asynchronous optimization is appealing as it offers 100% pipeline utilization by construction. However, it is inherently challenging as the weights and gradients are no longer synchronized, leading to stale (or delayed) gradients. To alleviate this, we introduce a variant of Nesterov Accelerated Gradient (NAG) for asynchronous optimization in PP. Specifically, we modify the look-ahead step in NAG to effectively address t","authors_text":"Alexander Long, Gil Avraham, Sameera Ramasinghe, Thalaiyasingam Ajanthan, Yan Zuo","cross_cats":["cs.DC"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-02T08:23:29Z","title":"Nesterov Method for Asynchronous Pipeline Parallel Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.01099","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":"sha256:06b27c4d89b0bbba6d2656c5b71150619105dc0b264eaae973c036dd11df942e","target":"record","created_at":"2026-07-05T10:57:45Z","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":"189ddfab2afc990fc9336a42e3fab6ecadc28d6fa22ba310864914d12586f6cc","cross_cats_sorted":["cs.DC"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-02T08:23:29Z","title_canon_sha256":"678cb8153bcb9c5a52d3c4f59365687d5f0684a2113729f7a78ca3b8e3ac86b7"},"schema_version":"1.0","source":{"id":"2505.01099","kind":"arxiv","version":1}},"canonical_sha256":"d37a4277a6697b337367935e3ba07a96b780a50c64df857010b27309570af48e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d37a4277a6697b337367935e3ba07a96b780a50c64df857010b27309570af48e","first_computed_at":"2026-07-05T10:57:45.107404Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:57:45.107404Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vl0GMiv7nmwbmuAxodL1A/mruUdIOwxdum2lkdAgnPVQX3HqaYNAIYV96ZE6S/aWJufncoBsSIgXSu2fmC6zBg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:57:45.107934Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.01099","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:06b27c4d89b0bbba6d2656c5b71150619105dc0b264eaae973c036dd11df942e","sha256:a10ee05072e6ff21c758b676b85f24e0d2f7e39cf5cf8a78de87568c5a7e8875"],"state_sha256":"bdd909dfe13652cf418f74def07259241701ccd939b858958816469558603819"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eSHH8G/8QQynUVPrspvXW1TxRx2bliaz1sysx6qlm+uD4nWnonuEPqKFxpJJFeuPoUAhYU3mUrHy+0A+7cCQAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-18T06:44:27.173026Z","bundle_sha256":"8ffe27274c99973c5d7d2d072aba4436c7efe667bc8401863616ae7b2f94a19f"}}