{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:7GPYI7W74ENF4LEP6QFTHOKCW5","short_pith_number":"pith:7GPYI7W7","canonical_record":{"source":{"id":"2502.04164","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T15:47:18Z","cross_cats_sorted":[],"title_canon_sha256":"16e68df10db98d683aeeb283cc82368b6f2b517e1e695ac716eefd8a92d612b5","abstract_canon_sha256":"a7d59045eb62f01f57e620750a07e56c6bf2f868b7ba01ddf0c9751c2fa1d56f"},"schema_version":"1.0"},"canonical_sha256":"f99f847edfe11a5e2c8ff40b33b942b752cca41031ae9847f9394e2e3168349a","source":{"kind":"arxiv","id":"2502.04164","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.04164","created_at":"2026-07-05T11:53:43Z"},{"alias_kind":"arxiv_version","alias_value":"2502.04164v2","created_at":"2026-07-05T11:53:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04164","created_at":"2026-07-05T11:53:43Z"},{"alias_kind":"pith_short_12","alias_value":"7GPYI7W74ENF","created_at":"2026-07-05T11:53:43Z"},{"alias_kind":"pith_short_16","alias_value":"7GPYI7W74ENF4LEP","created_at":"2026-07-05T11:53:43Z"},{"alias_kind":"pith_short_8","alias_value":"7GPYI7W7","created_at":"2026-07-05T11:53:43Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:7GPYI7W74ENF4LEP6QFTHOKCW5","target":"record","payload":{"canonical_record":{"source":{"id":"2502.04164","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T15:47:18Z","cross_cats_sorted":[],"title_canon_sha256":"16e68df10db98d683aeeb283cc82368b6f2b517e1e695ac716eefd8a92d612b5","abstract_canon_sha256":"a7d59045eb62f01f57e620750a07e56c6bf2f868b7ba01ddf0c9751c2fa1d56f"},"schema_version":"1.0"},"canonical_sha256":"f99f847edfe11a5e2c8ff40b33b942b752cca41031ae9847f9394e2e3168349a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:53:43.965659Z","signature_b64":"DRXqv+OydtZxEu4kiTB+n2lNxslVyRAp5Zf/LUvYmDFvRxj/EiUEVH616Z1FcIFDEBWQPZoSo3JeNz74qVbvBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f99f847edfe11a5e2c8ff40b33b942b752cca41031ae9847f9394e2e3168349a","last_reissued_at":"2026-07-05T11:53:43.965189Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:53:43.965189Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.04164","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:53:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0Ob5CZ39FibDqIwTPCFtJz5HyatpQ1aLnpN1Px03jlRKTmgIkBjk5poGvVg0OP2q2B7IRtbuvEcpP2/DXSAKCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T23:17:34.668960Z"},"content_sha256":"a13c3a73b2cad33bb4de31984be7b3d380f41d7919cca2d38a32c720b1eea657","schema_version":"1.0","event_id":"sha256:a13c3a73b2cad33bb4de31984be7b3d380f41d7919cca2d38a32c720b1eea657"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:7GPYI7W74ENF4LEP6QFTHOKCW5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Efficient Distributed Optimization under Heavy-Tailed Noise","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Manzil Zaheer, Su Hyeong Lee, Tian Li","submitted_at":"2025-02-06T15:47:18Z","abstract_excerpt":"Distributed optimization has become the default training paradigm in modern machine learning due to the growing scale of models and datasets. To mitigate communication overhead, local updates are often applied before global aggregation, resulting in a nested optimization approach with inner and outer steps. However, heavy-tailed stochastic gradient noise remains a significant challenge, particularly in attention-based models, hindering effective training. In this work, we propose TailOPT, an efficient framework designed to address heavy-tailed noise by leveraging adaptive optimization or clipp"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04164","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/2502.04164/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":"2026-07-05T11:53:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ag1hTOXsvBVVf/r9Cg9oxwEEXMFY8CGtbBlbM14KVELFrIbfqA0Z4YxR0Nilgny8Oe953jP7e3S+8q/0QYQ9Ag==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T23:17:34.670091Z"},"content_sha256":"3ad8d55d3aa7f1ce31699950eb9d92405af161805f9546e813ba40885cfcdc8c","schema_version":"1.0","event_id":"sha256:3ad8d55d3aa7f1ce31699950eb9d92405af161805f9546e813ba40885cfcdc8c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/7GPYI7W74ENF4LEP6QFTHOKCW5/bundle.json","state_url":"https://pith.science/pith/7GPYI7W74ENF4LEP6QFTHOKCW5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/7GPYI7W74ENF4LEP6QFTHOKCW5/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":"2026-08-04T23:17:34Z","links":{"resolver":"https://pith.science/pith/7GPYI7W74ENF4LEP6QFTHOKCW5","bundle":"https://pith.science/pith/7GPYI7W74ENF4LEP6QFTHOKCW5/bundle.json","state":"https://pith.science/pith/7GPYI7W74ENF4LEP6QFTHOKCW5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/7GPYI7W74ENF4LEP6QFTHOKCW5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:7GPYI7W74ENF4LEP6QFTHOKCW5","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":"a7d59045eb62f01f57e620750a07e56c6bf2f868b7ba01ddf0c9751c2fa1d56f","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T15:47:18Z","title_canon_sha256":"16e68df10db98d683aeeb283cc82368b6f2b517e1e695ac716eefd8a92d612b5"},"schema_version":"1.0","source":{"id":"2502.04164","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.04164","created_at":"2026-07-05T11:53:43Z"},{"alias_kind":"arxiv_version","alias_value":"2502.04164v2","created_at":"2026-07-05T11:53:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04164","created_at":"2026-07-05T11:53:43Z"},{"alias_kind":"pith_short_12","alias_value":"7GPYI7W74ENF","created_at":"2026-07-05T11:53:43Z"},{"alias_kind":"pith_short_16","alias_value":"7GPYI7W74ENF4LEP","created_at":"2026-07-05T11:53:43Z"},{"alias_kind":"pith_short_8","alias_value":"7GPYI7W7","created_at":"2026-07-05T11:53:43Z"}],"graph_snapshots":[{"event_id":"sha256:3ad8d55d3aa7f1ce31699950eb9d92405af161805f9546e813ba40885cfcdc8c","target":"graph","created_at":"2026-07-05T11:53:43Z","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/2502.04164/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Distributed optimization has become the default training paradigm in modern machine learning due to the growing scale of models and datasets. To mitigate communication overhead, local updates are often applied before global aggregation, resulting in a nested optimization approach with inner and outer steps. However, heavy-tailed stochastic gradient noise remains a significant challenge, particularly in attention-based models, hindering effective training. In this work, we propose TailOPT, an efficient framework designed to address heavy-tailed noise by leveraging adaptive optimization or clipp","authors_text":"Manzil Zaheer, Su Hyeong Lee, Tian Li","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T15:47:18Z","title":"Efficient Distributed Optimization under Heavy-Tailed Noise"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04164","kind":"arxiv","version":2},"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:a13c3a73b2cad33bb4de31984be7b3d380f41d7919cca2d38a32c720b1eea657","target":"record","created_at":"2026-07-05T11:53:43Z","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":"a7d59045eb62f01f57e620750a07e56c6bf2f868b7ba01ddf0c9751c2fa1d56f","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T15:47:18Z","title_canon_sha256":"16e68df10db98d683aeeb283cc82368b6f2b517e1e695ac716eefd8a92d612b5"},"schema_version":"1.0","source":{"id":"2502.04164","kind":"arxiv","version":2}},"canonical_sha256":"f99f847edfe11a5e2c8ff40b33b942b752cca41031ae9847f9394e2e3168349a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f99f847edfe11a5e2c8ff40b33b942b752cca41031ae9847f9394e2e3168349a","first_computed_at":"2026-07-05T11:53:43.965189Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:53:43.965189Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"DRXqv+OydtZxEu4kiTB+n2lNxslVyRAp5Zf/LUvYmDFvRxj/EiUEVH616Z1FcIFDEBWQPZoSo3JeNz74qVbvBA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:53:43.965659Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.04164","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a13c3a73b2cad33bb4de31984be7b3d380f41d7919cca2d38a32c720b1eea657","sha256:3ad8d55d3aa7f1ce31699950eb9d92405af161805f9546e813ba40885cfcdc8c"],"state_sha256":"af31cc706a2e0d77f7fbf1ba03cd80fafaac7a1e239770c3a9a8fa5ae99a2f66"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9mxNbvILi3NjzN0tdj+mDYz4cns2R1nsacsZFYB67HAqm4MYx4ABlyCOC8kn+PFU4H752y1wtVO3MtZE9Bk0Dw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T23:17:34.676695Z","bundle_sha256":"27550a5421232bd29f303cb41fa3a1d9c4d4418b61e1230df45eda31b3478c2f"}}