{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:MB47VM45YODGVZQEBYU7WAEM3T","short_pith_number":"pith:MB47VM45","canonical_record":{"source":{"id":"2605.13434","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T12:27:22Z","cross_cats_sorted":["cs.DC","math.OC","stat.ML"],"title_canon_sha256":"460975d702698bd8d60f688c91f2124e253de33f880bb3d8fe8c5ddd4ab55b90","abstract_canon_sha256":"013e29306d44dbd8adc4c101bf414a6403d2edaecc85e2292f97c5010778a9dd"},"schema_version":"1.0"},"canonical_sha256":"6079fab39dc3866ae6040e29fb008cdcddec03046b3a673109882c1096967fa9","source":{"kind":"arxiv","id":"2605.13434","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13434","created_at":"2026-05-18T02:44:47Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13434v1","created_at":"2026-05-18T02:44:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13434","created_at":"2026-05-18T02:44:47Z"},{"alias_kind":"pith_short_12","alias_value":"MB47VM45YODG","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"MB47VM45YODGVZQE","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"MB47VM45","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:MB47VM45YODGVZQEBYU7WAEM3T","target":"record","payload":{"canonical_record":{"source":{"id":"2605.13434","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T12:27:22Z","cross_cats_sorted":["cs.DC","math.OC","stat.ML"],"title_canon_sha256":"460975d702698bd8d60f688c91f2124e253de33f880bb3d8fe8c5ddd4ab55b90","abstract_canon_sha256":"013e29306d44dbd8adc4c101bf414a6403d2edaecc85e2292f97c5010778a9dd"},"schema_version":"1.0"},"canonical_sha256":"6079fab39dc3866ae6040e29fb008cdcddec03046b3a673109882c1096967fa9","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:44:47.132478Z","signature_b64":"pUp7KbyCqdpalci+3YAPUJZ8ty7c1kO6ScloBMQEfQgUO3ugbHnCkzVqyFskvzKjyPInQ+GsU0HPDXBkH/QxDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6079fab39dc3866ae6040e29fb008cdcddec03046b3a673109882c1096967fa9","last_reissued_at":"2026-05-18T02:44:47.131927Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:44:47.131927Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.13434","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":"2026-05-18T02:44:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jTDT/7wm5WU8Go1kPhemtyyW92jnq+qUhiUa7Xy+e495W2G01r9ylI4SEZUsFFl6axg1Y0XmCjGZp+dFuTVwBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T04:31:33.639672Z"},"content_sha256":"ce3d721d4291d45e9b453af89e1813ba0006d1760314207f0489b370e6e5444c","schema_version":"1.0","event_id":"sha256:ce3d721d4291d45e9b453af89e1813ba0006d1760314207f0489b370e6e5444c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:MB47VM45YODGVZQEBYU7WAEM3T","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Rescaled Asynchronous SGD: Optimal Distributed Optimization under Data and System Heterogeneity","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Rescaling worker stepsizes by computation time fixes bias in asynchronous SGD so it converges to the true global objective.","cross_cats":["cs.DC","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Ammar Mahran, Artavazd Maranjyan, Peter Richt\\'arik","submitted_at":"2026-05-13T12:27:22Z","abstract_excerpt":"Asynchronous stochastic gradient descent (ASGD) is a standard way to exploit heterogeneous compute resources in distributed learning: instead of forcing fast workers to wait for slow ones, the server updates the model whenever a gradient arrives. Vanilla ASGD applies each arriving gradient with the same weight. When local data distributions are heterogeneous, this becomes problematic: faster workers contribute more updates, and we show theoretically that the method is biased toward a frequency-weighted average of the local objectives rather than the desired global objective. Existing remedies "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"we prove that the resulting method, Rescaled ASGD, converges to stationary points of the correct global objective in the fixed-computation model. Its time complexity matches the known lower bound in the leading term, while the effects of staleness and data heterogeneity appear only in lower-order terms.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"under smoothness and bounded heterogeneity assumptions","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Rescaled ASGD recovers convergence to the true global objective by rescaling worker stepsizes proportional to computation times, matching the known time lower bound in the leading term under non-convex smoothness and bounded heterogeneity.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Rescaling worker stepsizes by computation time fixes bias in asynchronous SGD so it converges to the true global objective.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"bed43d911b513c3ba7be443a98374a6b8d93807a454477bcce4e5a28607203ab"},"source":{"id":"2605.13434","kind":"arxiv","version":1},"verdict":{"id":"3df6c769-5d61-42f4-b85b-e2d8b0175c8e","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:27:27.645909Z","strongest_claim":"we prove that the resulting method, Rescaled ASGD, converges to stationary points of the correct global objective in the fixed-computation model. Its time complexity matches the known lower bound in the leading term, while the effects of staleness and data heterogeneity appear only in lower-order terms.","one_line_summary":"Rescaled ASGD recovers convergence to the true global objective by rescaling worker stepsizes proportional to computation times, matching the known time lower bound in the leading term under non-convex smoothness and bounded heterogeneity.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"under smoothness and bounded heterogeneity assumptions","pith_extraction_headline":"Rescaling worker stepsizes by computation time fixes bias in asynchronous SGD so it converges to the true global objective."},"references":{"count":300,"sample":[{"doi":"10.1109/msp.2020.2975749","year":2020,"title":"Federated Learning: Chal- lenges, Methods, and Future Directions.IEEE Signal Processing Magazine, 37(3):50–60","work_id":"4916403b-b4d2-45a0-9053-7b8f137c878f","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.48550/arxiv.1909.06335","year":1909,"title":"Measuring the Effects of Non-Identical Data Distribution for Federated Visual Classification","work_id":"1e8d2981-bd90-4c86-8676-223498b6d816","ref_index":2,"cited_arxiv_id":"1909.06335","is_internal_anchor":true},{"doi":"10.1109/tit.2017.2736066","year":2015,"title":"arXiv.org , author =","work_id":"42d7abc2-f71a-4d63-885a-c432425462ef","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.48550/arxiv.1812.01097","year":null,"title":"Leaf: A benchmark for federated settings","work_id":"94ec8be1-6cf1-4962-994c-e762a0b9c3ee","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2012,"title":"arXiv.org , author =","work_id":"a670dcca-8345-441d-9855-f6bc678e0fbd","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":300,"snapshot_sha256":"421548bf02e07daa14540f2b965d273433a08dbfcc992d1b513da02c06f42655","internal_anchors":34},"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":"3df6c769-5d61-42f4-b85b-e2d8b0175c8e"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EYLWH4/2c/8KPn56AdcpNIFlzT8EyP7imTyJv3s+wJPCraG3RkG5I7GstspXuGBtxY6NimxJvmbPABkEHBknDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T04:31:33.640693Z"},"content_sha256":"f9c1cedae106587e7832edf361749502d52c84c886743f6b5dbcda443c79e12c","schema_version":"1.0","event_id":"sha256:f9c1cedae106587e7832edf361749502d52c84c886743f6b5dbcda443c79e12c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MB47VM45YODGVZQEBYU7WAEM3T/bundle.json","state_url":"https://pith.science/pith/MB47VM45YODGVZQEBYU7WAEM3T/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MB47VM45YODGVZQEBYU7WAEM3T/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":"2026-05-22T04:31:33Z","links":{"resolver":"https://pith.science/pith/MB47VM45YODGVZQEBYU7WAEM3T","bundle":"https://pith.science/pith/MB47VM45YODGVZQEBYU7WAEM3T/bundle.json","state":"https://pith.science/pith/MB47VM45YODGVZQEBYU7WAEM3T/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MB47VM45YODGVZQEBYU7WAEM3T/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:MB47VM45YODGVZQEBYU7WAEM3T","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":"013e29306d44dbd8adc4c101bf414a6403d2edaecc85e2292f97c5010778a9dd","cross_cats_sorted":["cs.DC","math.OC","stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T12:27:22Z","title_canon_sha256":"460975d702698bd8d60f688c91f2124e253de33f880bb3d8fe8c5ddd4ab55b90"},"schema_version":"1.0","source":{"id":"2605.13434","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13434","created_at":"2026-05-18T02:44:47Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13434v1","created_at":"2026-05-18T02:44:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13434","created_at":"2026-05-18T02:44:47Z"},{"alias_kind":"pith_short_12","alias_value":"MB47VM45YODG","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"MB47VM45YODGVZQE","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"MB47VM45","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:f9c1cedae106587e7832edf361749502d52c84c886743f6b5dbcda443c79e12c","target":"graph","created_at":"2026-05-18T02:44:47Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"we prove that the resulting method, Rescaled ASGD, converges to stationary points of the correct global objective in the fixed-computation model. Its time complexity matches the known lower bound in the leading term, while the effects of staleness and data heterogeneity appear only in lower-order terms."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"under smoothness and bounded heterogeneity assumptions"},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Rescaled ASGD recovers convergence to the true global objective by rescaling worker stepsizes proportional to computation times, matching the known time lower bound in the leading term under non-convex smoothness and bounded heterogeneity."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Rescaling worker stepsizes by computation time fixes bias in asynchronous SGD so it converges to the true global objective."}],"snapshot_sha256":"bed43d911b513c3ba7be443a98374a6b8d93807a454477bcce4e5a28607203ab"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Asynchronous stochastic gradient descent (ASGD) is a standard way to exploit heterogeneous compute resources in distributed learning: instead of forcing fast workers to wait for slow ones, the server updates the model whenever a gradient arrives. Vanilla ASGD applies each arriving gradient with the same weight. When local data distributions are heterogeneous, this becomes problematic: faster workers contribute more updates, and we show theoretically that the method is biased toward a frequency-weighted average of the local objectives rather than the desired global objective. Existing remedies ","authors_text":"Ammar Mahran, Artavazd Maranjyan, Peter Richt\\'arik","cross_cats":["cs.DC","math.OC","stat.ML"],"headline":"Rescaling worker stepsizes by computation time fixes bias in asynchronous SGD so it converges to the true global objective.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T12:27:22Z","title":"Rescaled Asynchronous SGD: Optimal Distributed Optimization under Data and System Heterogeneity"},"references":{"count":300,"internal_anchors":34,"resolved_work":300,"sample":[{"cited_arxiv_id":"","doi":"10.1109/msp.2020.2975749","is_internal_anchor":false,"ref_index":1,"title":"Federated Learning: Chal- lenges, Methods, and Future Directions.IEEE Signal Processing Magazine, 37(3):50–60","work_id":"4916403b-b4d2-45a0-9053-7b8f137c878f","year":2020},{"cited_arxiv_id":"1909.06335","doi":"10.48550/arxiv.1909.06335","is_internal_anchor":true,"ref_index":2,"title":"Measuring the Effects of Non-Identical Data Distribution for Federated Visual Classification","work_id":"1e8d2981-bd90-4c86-8676-223498b6d816","year":1909},{"cited_arxiv_id":"","doi":"10.1109/tit.2017.2736066","is_internal_anchor":false,"ref_index":3,"title":"arXiv.org , author =","work_id":"42d7abc2-f71a-4d63-885a-c432425462ef","year":2015},{"cited_arxiv_id":"","doi":"10.48550/arxiv.1812.01097","is_internal_anchor":false,"ref_index":4,"title":"Leaf: A benchmark for federated settings","work_id":"94ec8be1-6cf1-4962-994c-e762a0b9c3ee","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"arXiv.org , author =","work_id":"a670dcca-8345-441d-9855-f6bc678e0fbd","year":2012}],"snapshot_sha256":"421548bf02e07daa14540f2b965d273433a08dbfcc992d1b513da02c06f42655"},"source":{"id":"2605.13434","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T19:27:27.645909Z","id":"3df6c769-5d61-42f4-b85b-e2d8b0175c8e","model_set":{"reader":"grok-4.3"},"one_line_summary":"Rescaled ASGD recovers convergence to the true global objective by rescaling worker stepsizes proportional to computation times, matching the known time lower bound in the leading term under non-convex smoothness and bounded heterogeneity.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Rescaling worker stepsizes by computation time fixes bias in asynchronous SGD so it converges to the true global objective.","strongest_claim":"we prove that the resulting method, Rescaled ASGD, converges to stationary points of the correct global objective in the fixed-computation model. Its time complexity matches the known lower bound in the leading term, while the effects of staleness and data heterogeneity appear only in lower-order terms.","weakest_assumption":"under smoothness and bounded heterogeneity assumptions"}},"verdict_id":"3df6c769-5d61-42f4-b85b-e2d8b0175c8e"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:ce3d721d4291d45e9b453af89e1813ba0006d1760314207f0489b370e6e5444c","target":"record","created_at":"2026-05-18T02:44:47Z","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":"013e29306d44dbd8adc4c101bf414a6403d2edaecc85e2292f97c5010778a9dd","cross_cats_sorted":["cs.DC","math.OC","stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T12:27:22Z","title_canon_sha256":"460975d702698bd8d60f688c91f2124e253de33f880bb3d8fe8c5ddd4ab55b90"},"schema_version":"1.0","source":{"id":"2605.13434","kind":"arxiv","version":1}},"canonical_sha256":"6079fab39dc3866ae6040e29fb008cdcddec03046b3a673109882c1096967fa9","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6079fab39dc3866ae6040e29fb008cdcddec03046b3a673109882c1096967fa9","first_computed_at":"2026-05-18T02:44:47.131927Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:44:47.131927Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"pUp7KbyCqdpalci+3YAPUJZ8ty7c1kO6ScloBMQEfQgUO3ugbHnCkzVqyFskvzKjyPInQ+GsU0HPDXBkH/QxDQ==","signature_status":"signed_v1","signed_at":"2026-05-18T02:44:47.132478Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13434","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ce3d721d4291d45e9b453af89e1813ba0006d1760314207f0489b370e6e5444c","sha256:f9c1cedae106587e7832edf361749502d52c84c886743f6b5dbcda443c79e12c"],"state_sha256":"a1483848dc71058f0a3296608104db53ef7c0a4d8a14597812b6a4e4b346dc80"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OphSps71hHaIozrzu8gT0PlR+aZjOrPgzrh5j8IYitRJXqyvBkeQe0Nr6lngGHVEfudgyhNBXS+e/YqxHJrRBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-22T04:31:33.644716Z","bundle_sha256":"89bab41148d2dce2a27681740f8e783133585f7ca57582ad7e45ba048b1ab36d"}}