{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:STM76AJ3UGKFXMWTAWB6LKG54A","short_pith_number":"pith:STM76AJ3","canonical_record":{"source":{"id":"2605.27918","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-05-27T03:44:27Z","cross_cats_sorted":[],"title_canon_sha256":"6aff492b601dba14a52927ec235578949eb8a47cd17e94624546fae2d9663b93","abstract_canon_sha256":"7fee99768e20afe575921c5f48f6ae305203ec23e6db6b8b06f9c1e6bf59cf5a"},"schema_version":"1.0"},"canonical_sha256":"94d9ff013ba1945bb2d30583e5a8dde00eb80737cb145f2cc3d5f737211ac9f6","source":{"kind":"arxiv","id":"2605.27918","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.27918","created_at":"2026-05-28T01:04:52Z"},{"alias_kind":"arxiv_version","alias_value":"2605.27918v1","created_at":"2026-05-28T01:04:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.27918","created_at":"2026-05-28T01:04:52Z"},{"alias_kind":"pith_short_12","alias_value":"STM76AJ3UGKF","created_at":"2026-05-28T01:04:52Z"},{"alias_kind":"pith_short_16","alias_value":"STM76AJ3UGKFXMWT","created_at":"2026-05-28T01:04:52Z"},{"alias_kind":"pith_short_8","alias_value":"STM76AJ3","created_at":"2026-05-28T01:04:52Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:STM76AJ3UGKFXMWTAWB6LKG54A","target":"record","payload":{"canonical_record":{"source":{"id":"2605.27918","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-05-27T03:44:27Z","cross_cats_sorted":[],"title_canon_sha256":"6aff492b601dba14a52927ec235578949eb8a47cd17e94624546fae2d9663b93","abstract_canon_sha256":"7fee99768e20afe575921c5f48f6ae305203ec23e6db6b8b06f9c1e6bf59cf5a"},"schema_version":"1.0"},"canonical_sha256":"94d9ff013ba1945bb2d30583e5a8dde00eb80737cb145f2cc3d5f737211ac9f6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-28T01:04:52.513279Z","signature_b64":"gvx7u3ZqcV2XOgM8aTydGeevcUvN+ghZYNLsymnaoqie76Rx0MNOAe8eWWs9D+FW7atfk5BOlUglIrO2YBsHCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"94d9ff013ba1945bb2d30583e5a8dde00eb80737cb145f2cc3d5f737211ac9f6","last_reissued_at":"2026-05-28T01:04:52.512861Z","signature_status":"signed_v1","first_computed_at":"2026-05-28T01:04:52.512861Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.27918","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":"2026-05-28T01:04:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Bbe6FlTVu2Xq9FGrdE5askq4fIML3YGtpCPHNE27YtJLQ4xpGwN/bYH8OkaKJxDElTA9ZnBG5NPe5sGvUqQwBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-11T05:28:09.567523Z"},"content_sha256":"e33b24fb8bd397427aba5c463e88de52beabaa0ebe2778a0893e0f1e09d0c34c","schema_version":"1.0","event_id":"sha256:e33b24fb8bd397427aba5c463e88de52beabaa0ebe2778a0893e0f1e09d0c34c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:STM76AJ3UGKFXMWTAWB6LKG54A","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Addressing Variable Heterogeneity in Distributed Multimodal Training with Entrain","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Insu Jang, Mosharaf Chowdhury","submitted_at":"2026-05-27T03:44:27Z","abstract_excerpt":"Multimodal LLM datasets are inherently heterogeneous, with significant data variability. Although each modality exhibits independent variability, sample-level entanglement makes it difficult to balance workloads across both modalities and batches. We present Entrain, a distributed MLLM training framework that addresses both heterogeneity and variability in multimodal training workloads. Entrain challenges the intuition that dynamic data variability requires dynamic model parallelism by shifting the profiling paradigm from micro-level samples to macroscopic batches. We prove that a single, stat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.27918","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/2605.27918/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":"2026-05-28T01:04:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nS4dLHAuBHHFfFydPfo6JSmHayRwvn1piEH/FBHiD3hDXsnLtpZdD6Fg5c8l+FvU3YAKnWPCxaYj7r58JlN0CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-11T05:28:09.568265Z"},"content_sha256":"793e59adb13fa2b8f7c44a862ad5c7be05028da69960a78070c509e4619cf44d","schema_version":"1.0","event_id":"sha256:793e59adb13fa2b8f7c44a862ad5c7be05028da69960a78070c509e4619cf44d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/STM76AJ3UGKFXMWTAWB6LKG54A/bundle.json","state_url":"https://pith.science/pith/STM76AJ3UGKFXMWTAWB6LKG54A/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/STM76AJ3UGKFXMWTAWB6LKG54A/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":"2026-06-11T05:28:09Z","links":{"resolver":"https://pith.science/pith/STM76AJ3UGKFXMWTAWB6LKG54A","bundle":"https://pith.science/pith/STM76AJ3UGKFXMWTAWB6LKG54A/bundle.json","state":"https://pith.science/pith/STM76AJ3UGKFXMWTAWB6LKG54A/state.json","well_known_bundle":"https://pith.science/.well-known/pith/STM76AJ3UGKFXMWTAWB6LKG54A/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:STM76AJ3UGKFXMWTAWB6LKG54A","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":"7fee99768e20afe575921c5f48f6ae305203ec23e6db6b8b06f9c1e6bf59cf5a","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-05-27T03:44:27Z","title_canon_sha256":"6aff492b601dba14a52927ec235578949eb8a47cd17e94624546fae2d9663b93"},"schema_version":"1.0","source":{"id":"2605.27918","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.27918","created_at":"2026-05-28T01:04:52Z"},{"alias_kind":"arxiv_version","alias_value":"2605.27918v1","created_at":"2026-05-28T01:04:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.27918","created_at":"2026-05-28T01:04:52Z"},{"alias_kind":"pith_short_12","alias_value":"STM76AJ3UGKF","created_at":"2026-05-28T01:04:52Z"},{"alias_kind":"pith_short_16","alias_value":"STM76AJ3UGKFXMWT","created_at":"2026-05-28T01:04:52Z"},{"alias_kind":"pith_short_8","alias_value":"STM76AJ3","created_at":"2026-05-28T01:04:52Z"}],"graph_snapshots":[{"event_id":"sha256:793e59adb13fa2b8f7c44a862ad5c7be05028da69960a78070c509e4619cf44d","target":"graph","created_at":"2026-05-28T01:04:52Z","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/2605.27918/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multimodal LLM datasets are inherently heterogeneous, with significant data variability. Although each modality exhibits independent variability, sample-level entanglement makes it difficult to balance workloads across both modalities and batches. We present Entrain, a distributed MLLM training framework that addresses both heterogeneity and variability in multimodal training workloads. Entrain challenges the intuition that dynamic data variability requires dynamic model parallelism by shifting the profiling paradigm from micro-level samples to macroscopic batches. We prove that a single, stat","authors_text":"Insu Jang, Mosharaf Chowdhury","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-05-27T03:44:27Z","title":"Addressing Variable Heterogeneity in Distributed Multimodal Training with Entrain"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.27918","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:e33b24fb8bd397427aba5c463e88de52beabaa0ebe2778a0893e0f1e09d0c34c","target":"record","created_at":"2026-05-28T01:04:52Z","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":"7fee99768e20afe575921c5f48f6ae305203ec23e6db6b8b06f9c1e6bf59cf5a","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-05-27T03:44:27Z","title_canon_sha256":"6aff492b601dba14a52927ec235578949eb8a47cd17e94624546fae2d9663b93"},"schema_version":"1.0","source":{"id":"2605.27918","kind":"arxiv","version":1}},"canonical_sha256":"94d9ff013ba1945bb2d30583e5a8dde00eb80737cb145f2cc3d5f737211ac9f6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"94d9ff013ba1945bb2d30583e5a8dde00eb80737cb145f2cc3d5f737211ac9f6","first_computed_at":"2026-05-28T01:04:52.512861Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-28T01:04:52.512861Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"gvx7u3ZqcV2XOgM8aTydGeevcUvN+ghZYNLsymnaoqie76Rx0MNOAe8eWWs9D+FW7atfk5BOlUglIrO2YBsHCg==","signature_status":"signed_v1","signed_at":"2026-05-28T01:04:52.513279Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.27918","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e33b24fb8bd397427aba5c463e88de52beabaa0ebe2778a0893e0f1e09d0c34c","sha256:793e59adb13fa2b8f7c44a862ad5c7be05028da69960a78070c509e4619cf44d"],"state_sha256":"4d053cff962e1d28c66b3ac8e1b1ff5126456c36b8b8c0ca98c8bc1849f9563e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TYL+kKJaTY3PL7CJkLpRMUmBUxBgWkCBzx+CA1B34Fm2ciZFukfg/9Lsexv/i5mFz8Q4L05hC6bCZZWfAY2lBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-11T05:28:09.572907Z","bundle_sha256":"e01e3da69e7dfab2408bacfbbaf85c26c8fb066eed3a70804ee5340af1e1bdc5"}}