{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:IUO2D37CHJK57UHKMP5LSFFOEP","short_pith_number":"pith:IUO2D37C","schema_version":"1.0","canonical_sha256":"451da1efe23a55dfd0ea63fab914ae23d5fece5866df24be4cdc5a2b56f12740","source":{"kind":"arxiv","id":"2607.01678","version":1},"attestation_state":"computed","paper":{"title":"SCAPE: Accurate and Efficient LLM Training with Extreme Sparse Communication","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Haotian Xie, Junlin Chen, Mingkai Zheng, Zhao Zhang","submitted_at":"2026-07-02T04:10:42Z","abstract_excerpt":"Communication increasingly dominates the cost of Large Language Model (LLM) pre-training, especially under data-parallel and sharded training schemes, where gradient synchronization and parameter reconstruction overhead increase with model size and system scale. Existing communication-reduction methods either sparsify raw gradients, which can be unstable for modern Adam-style optimizers at high sparsity, or quantize communication, whose savings are fundamentally bounded by bit width and often incur additional runtime overhead. We present SCAPE, a communication-efficient distributed optimizer f"},"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":"2607.01678","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-02T04:10:42Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"b34b5a91c0656b4def634b59e156d4e419efd07d5a1c0d87ed7a3442fa6bddd5","abstract_canon_sha256":"39b62244bd74cebc5311721e075650510ca74534ee76e0fca0f5272eaaa81469"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-03T01:17:26.555380Z","signature_b64":"3IcaSpxd11EO7S7fcddS+HZl1kHw+rGWmyZ9e5aW0z7yL4cwXVFH7xfjx+4NIDBeUSbOM82GbRmJuOqGdxe7CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"451da1efe23a55dfd0ea63fab914ae23d5fece5866df24be4cdc5a2b56f12740","last_reissued_at":"2026-07-03T01:17:26.554885Z","signature_status":"signed_v1","first_computed_at":"2026-07-03T01:17:26.554885Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SCAPE: Accurate and Efficient LLM Training with Extreme Sparse Communication","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Haotian Xie, Junlin Chen, Mingkai Zheng, Zhao Zhang","submitted_at":"2026-07-02T04:10:42Z","abstract_excerpt":"Communication increasingly dominates the cost of Large Language Model (LLM) pre-training, especially under data-parallel and sharded training schemes, where gradient synchronization and parameter reconstruction overhead increase with model size and system scale. Existing communication-reduction methods either sparsify raw gradients, which can be unstable for modern Adam-style optimizers at high sparsity, or quantize communication, whose savings are fundamentally bounded by bit width and often incur additional runtime overhead. We present SCAPE, a communication-efficient distributed optimizer f"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.01678","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/2607.01678/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":"2607.01678","created_at":"2026-07-03T01:17:26.554948+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.01678v1","created_at":"2026-07-03T01:17:26.554948+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.01678","created_at":"2026-07-03T01:17:26.554948+00:00"},{"alias_kind":"pith_short_12","alias_value":"IUO2D37CHJK5","created_at":"2026-07-03T01:17:26.554948+00:00"},{"alias_kind":"pith_short_16","alias_value":"IUO2D37CHJK57UHK","created_at":"2026-07-03T01:17:26.554948+00:00"},{"alias_kind":"pith_short_8","alias_value":"IUO2D37C","created_at":"2026-07-03T01:17:26.554948+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IUO2D37CHJK57UHKMP5LSFFOEP","json":"https://pith.science/pith/IUO2D37CHJK57UHKMP5LSFFOEP.json","graph_json":"https://pith.science/api/pith-number/IUO2D37CHJK57UHKMP5LSFFOEP/graph.json","events_json":"https://pith.science/api/pith-number/IUO2D37CHJK57UHKMP5LSFFOEP/events.json","paper":"https://pith.science/paper/IUO2D37C"},"agent_actions":{"view_html":"https://pith.science/pith/IUO2D37CHJK57UHKMP5LSFFOEP","download_json":"https://pith.science/pith/IUO2D37CHJK57UHKMP5LSFFOEP.json","view_paper":"https://pith.science/paper/IUO2D37C","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.01678&json=true","fetch_graph":"https://pith.science/api/pith-number/IUO2D37CHJK57UHKMP5LSFFOEP/graph.json","fetch_events":"https://pith.science/api/pith-number/IUO2D37CHJK57UHKMP5LSFFOEP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IUO2D37CHJK57UHKMP5LSFFOEP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IUO2D37CHJK57UHKMP5LSFFOEP/action/storage_attestation","attest_author":"https://pith.science/pith/IUO2D37CHJK57UHKMP5LSFFOEP/action/author_attestation","sign_citation":"https://pith.science/pith/IUO2D37CHJK57UHKMP5LSFFOEP/action/citation_signature","submit_replication":"https://pith.science/pith/IUO2D37CHJK57UHKMP5LSFFOEP/action/replication_record"}},"created_at":"2026-07-03T01:17:26.554948+00:00","updated_at":"2026-07-03T01:17:26.554948+00:00"}