{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:JIQOTZ7JYZAS6XNH7VFAP3RK5P","short_pith_number":"pith:JIQOTZ7J","canonical_record":{"source":{"id":"1901.09269","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-26T19:50:45Z","cross_cats_sorted":["math.OC","stat.ML"],"title_canon_sha256":"3f34e856ce9a9ea20c9f9131c37b55a5b0573136eecaae1f2bf7fc8cd822bae5","abstract_canon_sha256":"8582d49ad493ac035006c07ecb0a9840d081a8beedddf1437bdb87052b7577a9"},"schema_version":"1.0"},"canonical_sha256":"4a20e9e7e9c6412f5da7fd4a07ee2aebeb064042fbfb6f07e374a1bc2c95a21f","source":{"kind":"arxiv","id":"1901.09269","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1901.09269","created_at":"2026-07-05T07:28:24Z"},{"alias_kind":"arxiv_version","alias_value":"1901.09269v3","created_at":"2026-07-05T07:28:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1901.09269","created_at":"2026-07-05T07:28:24Z"},{"alias_kind":"pith_short_12","alias_value":"JIQOTZ7JYZAS","created_at":"2026-07-05T07:28:24Z"},{"alias_kind":"pith_short_16","alias_value":"JIQOTZ7JYZAS6XNH","created_at":"2026-07-05T07:28:24Z"},{"alias_kind":"pith_short_8","alias_value":"JIQOTZ7J","created_at":"2026-07-05T07:28:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:JIQOTZ7JYZAS6XNH7VFAP3RK5P","target":"record","payload":{"canonical_record":{"source":{"id":"1901.09269","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-26T19:50:45Z","cross_cats_sorted":["math.OC","stat.ML"],"title_canon_sha256":"3f34e856ce9a9ea20c9f9131c37b55a5b0573136eecaae1f2bf7fc8cd822bae5","abstract_canon_sha256":"8582d49ad493ac035006c07ecb0a9840d081a8beedddf1437bdb87052b7577a9"},"schema_version":"1.0"},"canonical_sha256":"4a20e9e7e9c6412f5da7fd4a07ee2aebeb064042fbfb6f07e374a1bc2c95a21f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:28:24.389103Z","signature_b64":"it8mvNX7vhmS/O6ddDcTUqAwAWlDkkjaMKBg8yOSs6Ww7RquOXLBu+gqYWa1X+elLFi05xKOcjOrNUqMNDHRCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4a20e9e7e9c6412f5da7fd4a07ee2aebeb064042fbfb6f07e374a1bc2c95a21f","last_reissued_at":"2026-07-05T07:28:24.388631Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:28:24.388631Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1901.09269","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:28:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"krSbarhrE+AIASAUKyIS7sfZb7L86jMLqhMDWvyxZQMq3VGE8jH3wcd+jokNwRNcVJQUIqvIskdC+pXh7vEAAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T15:38:30.888210Z"},"content_sha256":"f3a086c1fe63a65e651c6b9dff416b9eba62e33ff03a28bfbf39bce9e46693e5","schema_version":"1.0","event_id":"sha256:f3a086c1fe63a65e651c6b9dff416b9eba62e33ff03a28bfbf39bce9e46693e5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:JIQOTZ7JYZAS6XNH7VFAP3RK5P","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Distributed Learning with Compressed Gradient Differences","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Eduard Gorbunov, Konstantin Mishchenko, Martin Tak\\'a\\v{c}, Peter Richt\\'arik","submitted_at":"2019-01-26T19:50:45Z","abstract_excerpt":"Training large machine learning models requires a distributed computing approach, with communication of the model updates being the bottleneck. For this reason, several methods based on the compression (e.g., sparsification and/or quantization) of updates were recently proposed, including QSGD (Alistarh et al., 2017), TernGrad (Wen et al., 2017), SignSGD (Bernstein et al., 2018), and DQGD (Khirirat et al., 2018). However, none of these methods are able to learn the gradients, which renders them incapable of converging to the true optimum in the batch mode. In this work we propose a new distrib"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1901.09269","kind":"arxiv","version":3},"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/1901.09269/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":"2026-07-05T07:28:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"up5on0cYTXsNCcx0JhiEvByieVqsxW4LjxyBW0qXmD1ZI/h5TDKNJUmtorGbOdVDhvS+gHGjfnuv6NlQnnpFCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T15:38:30.888809Z"},"content_sha256":"80448911e6558bd07ca408d495213f906f155820ce3eb439721b9581d1eaac34","schema_version":"1.0","event_id":"sha256:80448911e6558bd07ca408d495213f906f155820ce3eb439721b9581d1eaac34"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JIQOTZ7JYZAS6XNH7VFAP3RK5P/bundle.json","state_url":"https://pith.science/pith/JIQOTZ7JYZAS6XNH7VFAP3RK5P/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JIQOTZ7JYZAS6XNH7VFAP3RK5P/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":"2026-08-04T15:38:30Z","links":{"resolver":"https://pith.science/pith/JIQOTZ7JYZAS6XNH7VFAP3RK5P","bundle":"https://pith.science/pith/JIQOTZ7JYZAS6XNH7VFAP3RK5P/bundle.json","state":"https://pith.science/pith/JIQOTZ7JYZAS6XNH7VFAP3RK5P/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JIQOTZ7JYZAS6XNH7VFAP3RK5P/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:JIQOTZ7JYZAS6XNH7VFAP3RK5P","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":"8582d49ad493ac035006c07ecb0a9840d081a8beedddf1437bdb87052b7577a9","cross_cats_sorted":["math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-26T19:50:45Z","title_canon_sha256":"3f34e856ce9a9ea20c9f9131c37b55a5b0573136eecaae1f2bf7fc8cd822bae5"},"schema_version":"1.0","source":{"id":"1901.09269","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1901.09269","created_at":"2026-07-05T07:28:24Z"},{"alias_kind":"arxiv_version","alias_value":"1901.09269v3","created_at":"2026-07-05T07:28:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1901.09269","created_at":"2026-07-05T07:28:24Z"},{"alias_kind":"pith_short_12","alias_value":"JIQOTZ7JYZAS","created_at":"2026-07-05T07:28:24Z"},{"alias_kind":"pith_short_16","alias_value":"JIQOTZ7JYZAS6XNH","created_at":"2026-07-05T07:28:24Z"},{"alias_kind":"pith_short_8","alias_value":"JIQOTZ7J","created_at":"2026-07-05T07:28:24Z"}],"graph_snapshots":[{"event_id":"sha256:80448911e6558bd07ca408d495213f906f155820ce3eb439721b9581d1eaac34","target":"graph","created_at":"2026-07-05T07:28:24Z","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/1901.09269/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Training large machine learning models requires a distributed computing approach, with communication of the model updates being the bottleneck. For this reason, several methods based on the compression (e.g., sparsification and/or quantization) of updates were recently proposed, including QSGD (Alistarh et al., 2017), TernGrad (Wen et al., 2017), SignSGD (Bernstein et al., 2018), and DQGD (Khirirat et al., 2018). However, none of these methods are able to learn the gradients, which renders them incapable of converging to the true optimum in the batch mode. In this work we propose a new distrib","authors_text":"Eduard Gorbunov, Konstantin Mishchenko, Martin Tak\\'a\\v{c}, Peter Richt\\'arik","cross_cats":["math.OC","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-26T19:50:45Z","title":"Distributed Learning with Compressed Gradient Differences"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1901.09269","kind":"arxiv","version":3},"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:f3a086c1fe63a65e651c6b9dff416b9eba62e33ff03a28bfbf39bce9e46693e5","target":"record","created_at":"2026-07-05T07:28:24Z","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":"8582d49ad493ac035006c07ecb0a9840d081a8beedddf1437bdb87052b7577a9","cross_cats_sorted":["math.OC","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-01-26T19:50:45Z","title_canon_sha256":"3f34e856ce9a9ea20c9f9131c37b55a5b0573136eecaae1f2bf7fc8cd822bae5"},"schema_version":"1.0","source":{"id":"1901.09269","kind":"arxiv","version":3}},"canonical_sha256":"4a20e9e7e9c6412f5da7fd4a07ee2aebeb064042fbfb6f07e374a1bc2c95a21f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4a20e9e7e9c6412f5da7fd4a07ee2aebeb064042fbfb6f07e374a1bc2c95a21f","first_computed_at":"2026-07-05T07:28:24.388631Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:28:24.388631Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"it8mvNX7vhmS/O6ddDcTUqAwAWlDkkjaMKBg8yOSs6Ww7RquOXLBu+gqYWa1X+elLFi05xKOcjOrNUqMNDHRCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T07:28:24.389103Z","signed_message":"canonical_sha256_bytes"},"source_id":"1901.09269","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f3a086c1fe63a65e651c6b9dff416b9eba62e33ff03a28bfbf39bce9e46693e5","sha256:80448911e6558bd07ca408d495213f906f155820ce3eb439721b9581d1eaac34"],"state_sha256":"a6b617eeff2e0b44f9894be3710afe5461e66a8e71d53544c1828cba087061b9"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zmTLxuRvaRKbYTCuwvnNXeilsOYKsa0rF2ZLFePSH/EzoRK3Nz2hAXxqPWhSzDzkBYk1ZPyqJ3wySz+h0WtcAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T15:38:30.920852Z","bundle_sha256":"7294d20d4b7ad7d7f8bc9f629dfb2e9d62522e18c71127f33610d8696608b30c"}}