{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:NB6NEJ2ELIO4R7C4INR6LOCLU5","short_pith_number":"pith:NB6NEJ2E","canonical_record":{"source":{"id":"2405.13226","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-21T22:26:01Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"b24ae81f10ba21b6c75ec72e267f6ba6342afa16eb9eec605f26732bad3620e4","abstract_canon_sha256":"c8a84a69cd26d8a8b6ce53af3bc18f47be8d7456b10e69b668f5f31a23662c49"},"schema_version":"1.0"},"canonical_sha256":"687cd227445a1dc8fc5c4363e5b84ba7668ffae009063d4564d6d7e00e5f19c4","source":{"kind":"arxiv","id":"2405.13226","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.13226","created_at":"2026-07-05T09:57:42Z"},{"alias_kind":"arxiv_version","alias_value":"2405.13226v2","created_at":"2026-07-05T09:57:42Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.13226","created_at":"2026-07-05T09:57:42Z"},{"alias_kind":"pith_short_12","alias_value":"NB6NEJ2ELIO4","created_at":"2026-07-05T09:57:42Z"},{"alias_kind":"pith_short_16","alias_value":"NB6NEJ2ELIO4R7C4","created_at":"2026-07-05T09:57:42Z"},{"alias_kind":"pith_short_8","alias_value":"NB6NEJ2E","created_at":"2026-07-05T09:57:42Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:NB6NEJ2ELIO4R7C4INR6LOCLU5","target":"record","payload":{"canonical_record":{"source":{"id":"2405.13226","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-21T22:26:01Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"b24ae81f10ba21b6c75ec72e267f6ba6342afa16eb9eec605f26732bad3620e4","abstract_canon_sha256":"c8a84a69cd26d8a8b6ce53af3bc18f47be8d7456b10e69b668f5f31a23662c49"},"schema_version":"1.0"},"canonical_sha256":"687cd227445a1dc8fc5c4363e5b84ba7668ffae009063d4564d6d7e00e5f19c4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:57:42.833727Z","signature_b64":"UBbNAgyMAwGTtPWGQcnFkvPGPRzbFq+qVd7KvfHnt+Kgkc7eLlFvIhC93I2wdPuOGM8NJQhFBq/MT+VhHK7vBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"687cd227445a1dc8fc5c4363e5b84ba7668ffae009063d4564d6d7e00e5f19c4","last_reissued_at":"2026-07-05T09:57:42.833219Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:57:42.833219Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2405.13226","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":"2026-07-05T09:57:42Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XDETQLJ2D92XMe7FDWKi4wYaE2U5F6q2vOgxhASVrpV9jg86/R3DB0/kd915HZA1zZfrKILCvcp4PCb7CjlkCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T15:48:20.175260Z"},"content_sha256":"ebc619fc95a6bc8c352af50ed62576281cc8bf3ef7ec18b209faf075e4271483","schema_version":"1.0","event_id":"sha256:ebc619fc95a6bc8c352af50ed62576281cc8bf3ef7ec18b209faf075e4271483"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:NB6NEJ2ELIO4R7C4INR6LOCLU5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Dataset Decomposition: Faster LLM Training with Variable Sequence Length Curriculum","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Cem Koc, Chun-Liang Li, Hadi Pouransari, Jen-Hao Rick Chang, Oncel Tuzel, Pavan Kumar Anasosalu Vasu, Vaishaal Shankar","submitted_at":"2024-05-21T22:26:01Z","abstract_excerpt":"Large language models (LLMs) are commonly trained on datasets consisting of fixed-length token sequences. These datasets are created by randomly concatenating documents of various lengths and then chunking them into sequences of a predetermined target length (concat-and-chunk). Recent attention implementations mask cross-document attention, reducing the effective length of a chunk of tokens. Additionally, training on long sequences becomes computationally prohibitive due to the quadratic cost of attention. In this study, we introduce dataset decomposition, a novel variable sequence length trai"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.13226","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/2405.13226/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":"2026-07-05T09:57:42Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QCK8TlzafBFs6gz1dsC5i/jf2dup9YEd8ypjQFW515o/mLOZ/+LdO1f/wbHZXi0ZU79kl97klWHGvCS6oTIRDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T15:48:20.175648Z"},"content_sha256":"f83f29a9fc492f4944adafe16fffda1d444f310048ae83d759b24e3bbeddcd6d","schema_version":"1.0","event_id":"sha256:f83f29a9fc492f4944adafe16fffda1d444f310048ae83d759b24e3bbeddcd6d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NB6NEJ2ELIO4R7C4INR6LOCLU5/bundle.json","state_url":"https://pith.science/pith/NB6NEJ2ELIO4R7C4INR6LOCLU5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NB6NEJ2ELIO4R7C4INR6LOCLU5/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":"2026-07-06T15:48:20Z","links":{"resolver":"https://pith.science/pith/NB6NEJ2ELIO4R7C4INR6LOCLU5","bundle":"https://pith.science/pith/NB6NEJ2ELIO4R7C4INR6LOCLU5/bundle.json","state":"https://pith.science/pith/NB6NEJ2ELIO4R7C4INR6LOCLU5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NB6NEJ2ELIO4R7C4INR6LOCLU5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:NB6NEJ2ELIO4R7C4INR6LOCLU5","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":"c8a84a69cd26d8a8b6ce53af3bc18f47be8d7456b10e69b668f5f31a23662c49","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-21T22:26:01Z","title_canon_sha256":"b24ae81f10ba21b6c75ec72e267f6ba6342afa16eb9eec605f26732bad3620e4"},"schema_version":"1.0","source":{"id":"2405.13226","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.13226","created_at":"2026-07-05T09:57:42Z"},{"alias_kind":"arxiv_version","alias_value":"2405.13226v2","created_at":"2026-07-05T09:57:42Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.13226","created_at":"2026-07-05T09:57:42Z"},{"alias_kind":"pith_short_12","alias_value":"NB6NEJ2ELIO4","created_at":"2026-07-05T09:57:42Z"},{"alias_kind":"pith_short_16","alias_value":"NB6NEJ2ELIO4R7C4","created_at":"2026-07-05T09:57:42Z"},{"alias_kind":"pith_short_8","alias_value":"NB6NEJ2E","created_at":"2026-07-05T09:57:42Z"}],"graph_snapshots":[{"event_id":"sha256:f83f29a9fc492f4944adafe16fffda1d444f310048ae83d759b24e3bbeddcd6d","target":"graph","created_at":"2026-07-05T09:57:42Z","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/2405.13226/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) are commonly trained on datasets consisting of fixed-length token sequences. These datasets are created by randomly concatenating documents of various lengths and then chunking them into sequences of a predetermined target length (concat-and-chunk). Recent attention implementations mask cross-document attention, reducing the effective length of a chunk of tokens. Additionally, training on long sequences becomes computationally prohibitive due to the quadratic cost of attention. In this study, we introduce dataset decomposition, a novel variable sequence length trai","authors_text":"Cem Koc, Chun-Liang Li, Hadi Pouransari, Jen-Hao Rick Chang, Oncel Tuzel, Pavan Kumar Anasosalu Vasu, Vaishaal Shankar","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-21T22:26:01Z","title":"Dataset Decomposition: Faster LLM Training with Variable Sequence Length Curriculum"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.13226","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:ebc619fc95a6bc8c352af50ed62576281cc8bf3ef7ec18b209faf075e4271483","target":"record","created_at":"2026-07-05T09:57:42Z","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":"c8a84a69cd26d8a8b6ce53af3bc18f47be8d7456b10e69b668f5f31a23662c49","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-21T22:26:01Z","title_canon_sha256":"b24ae81f10ba21b6c75ec72e267f6ba6342afa16eb9eec605f26732bad3620e4"},"schema_version":"1.0","source":{"id":"2405.13226","kind":"arxiv","version":2}},"canonical_sha256":"687cd227445a1dc8fc5c4363e5b84ba7668ffae009063d4564d6d7e00e5f19c4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"687cd227445a1dc8fc5c4363e5b84ba7668ffae009063d4564d6d7e00e5f19c4","first_computed_at":"2026-07-05T09:57:42.833219Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:57:42.833219Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"UBbNAgyMAwGTtPWGQcnFkvPGPRzbFq+qVd7KvfHnt+Kgkc7eLlFvIhC93I2wdPuOGM8NJQhFBq/MT+VhHK7vBg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:57:42.833727Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.13226","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ebc619fc95a6bc8c352af50ed62576281cc8bf3ef7ec18b209faf075e4271483","sha256:f83f29a9fc492f4944adafe16fffda1d444f310048ae83d759b24e3bbeddcd6d"],"state_sha256":"7e8b3e10133547aa9ea2a146856afa9208713acb0c8de3dd100c502c9482fb64"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KoTHz59SO9GJiCE2+Mp6e9pz3qMgVNtGUYEOOgI8VmdQCYZOS7vmW1wmDwVYXZa3mYqDOFGgH/z7YMx+aaApBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T15:48:20.177791Z","bundle_sha256":"7711ca67a2132a574fea0245d22fc0d07157de1ad9389f8e03bed42dcbd64499"}}