{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SVQDZ3RWBF6P5ZMVXFHBHKKMRE","short_pith_number":"pith:SVQDZ3RW","schema_version":"1.0","canonical_sha256":"95603cee36097cfee595b94e13a94c89009bd31445b446107b446297746de5c6","source":{"kind":"arxiv","id":"2505.12815","version":2},"attestation_state":"computed","paper":{"title":"Learning In Chaos: Efficient Autoscaling and Self-Healing for Multi-Party Distributed Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.DC","authors_text":"Gang Sun, Hongfang Yu, Long Luo, Mohsen Guizani, Qirong Ho, Rongxing Xiao, Steve Liu, Wenjiao Feng, Zonghang Li","submitted_at":"2025-05-19T07:52:17Z","abstract_excerpt":"Node and link churn in multi-party, cross-region clusters over wide-area networks (WANs) often disrupts distributed training. However, checkpoint-based recovery and cloud-centric autoscaling react slowly and assume centralized control, which is misaligned with the self-governed setup where institutions can freely join and leave. This paper proposes Chaos, a multi-party distributed training system with self-healing and autoscaling, enabling robust and elastic training under churn. It speeds up autoscaling via multi-neighbor state replication and model sharding. We formalize the sharding and ass"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2505.12815","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2025-05-19T07:52:17Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d1f3c90fd53f02e15d77c57af1c41356f1ee50ced82f399d0e7edfd762709f9b","abstract_canon_sha256":"70211e7945f4ee7dff53fde7f2c5d47973861633ff482690ed8a85155b41e757"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:11:17.977281Z","signature_b64":"WKX0WYU8TQBxlvRZ650CuyDQS6Ysdn+gzFcoiB9KqmmEHVx7FyfVNEkUeWQ38p87f/FFpVk2lx5QJOeHBBI8AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"95603cee36097cfee595b94e13a94c89009bd31445b446107b446297746de5c6","last_reissued_at":"2026-07-05T12:11:17.976725Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:11:17.976725Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning In Chaos: Efficient Autoscaling and Self-Healing for Multi-Party Distributed Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.DC","authors_text":"Gang Sun, Hongfang Yu, Long Luo, Mohsen Guizani, Qirong Ho, Rongxing Xiao, Steve Liu, Wenjiao Feng, Zonghang Li","submitted_at":"2025-05-19T07:52:17Z","abstract_excerpt":"Node and link churn in multi-party, cross-region clusters over wide-area networks (WANs) often disrupts distributed training. However, checkpoint-based recovery and cloud-centric autoscaling react slowly and assume centralized control, which is misaligned with the self-governed setup where institutions can freely join and leave. This paper proposes Chaos, a multi-party distributed training system with self-healing and autoscaling, enabling robust and elastic training under churn. It speeds up autoscaling via multi-neighbor state replication and model sharding. We formalize the sharding and ass"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12815","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/2505.12815/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2505.12815","created_at":"2026-07-05T12:11:17.976781+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.12815v2","created_at":"2026-07-05T12:11:17.976781+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12815","created_at":"2026-07-05T12:11:17.976781+00:00"},{"alias_kind":"pith_short_12","alias_value":"SVQDZ3RWBF6P","created_at":"2026-07-05T12:11:17.976781+00:00"},{"alias_kind":"pith_short_16","alias_value":"SVQDZ3RWBF6P5ZMV","created_at":"2026-07-05T12:11:17.976781+00:00"},{"alias_kind":"pith_short_8","alias_value":"SVQDZ3RW","created_at":"2026-07-05T12:11:17.976781+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.00672","citing_title":"Toward Edge General Intelligence with Multiple-Large Language Model (Multi-LLM): Architecture, Trust, and Orchestration","ref_index":77,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SVQDZ3RWBF6P5ZMVXFHBHKKMRE","json":"https://pith.science/pith/SVQDZ3RWBF6P5ZMVXFHBHKKMRE.json","graph_json":"https://pith.science/api/pith-number/SVQDZ3RWBF6P5ZMVXFHBHKKMRE/graph.json","events_json":"https://pith.science/api/pith-number/SVQDZ3RWBF6P5ZMVXFHBHKKMRE/events.json","paper":"https://pith.science/paper/SVQDZ3RW"},"agent_actions":{"view_html":"https://pith.science/pith/SVQDZ3RWBF6P5ZMVXFHBHKKMRE","download_json":"https://pith.science/pith/SVQDZ3RWBF6P5ZMVXFHBHKKMRE.json","view_paper":"https://pith.science/paper/SVQDZ3RW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.12815&json=true","fetch_graph":"https://pith.science/api/pith-number/SVQDZ3RWBF6P5ZMVXFHBHKKMRE/graph.json","fetch_events":"https://pith.science/api/pith-number/SVQDZ3RWBF6P5ZMVXFHBHKKMRE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SVQDZ3RWBF6P5ZMVXFHBHKKMRE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SVQDZ3RWBF6P5ZMVXFHBHKKMRE/action/storage_attestation","attest_author":"https://pith.science/pith/SVQDZ3RWBF6P5ZMVXFHBHKKMRE/action/author_attestation","sign_citation":"https://pith.science/pith/SVQDZ3RWBF6P5ZMVXFHBHKKMRE/action/citation_signature","submit_replication":"https://pith.science/pith/SVQDZ3RWBF6P5ZMVXFHBHKKMRE/action/replication_record"}},"created_at":"2026-07-05T12:11:17.976781+00:00","updated_at":"2026-07-05T12:11:17.976781+00:00"}