{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:QKFYFXODZU3WFDASUSTKANL4QY","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":"7b316213f566d638a7753779aa2379a393ac9a132b04e258c0fafc07ec7cb652","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T03:09:50Z","title_canon_sha256":"358fb531365be932999f0ea161178ce250f8ee0cbb25b031c7930c433a14d61f"},"schema_version":"1.0","source":{"id":"2605.17787","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.17787","created_at":"2026-05-20T00:04:58Z"},{"alias_kind":"arxiv_version","alias_value":"2605.17787v1","created_at":"2026-05-20T00:04:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17787","created_at":"2026-05-20T00:04:58Z"},{"alias_kind":"pith_short_12","alias_value":"QKFYFXODZU3W","created_at":"2026-05-20T00:04:58Z"},{"alias_kind":"pith_short_16","alias_value":"QKFYFXODZU3WFDAS","created_at":"2026-05-20T00:04:58Z"},{"alias_kind":"pith_short_8","alias_value":"QKFYFXOD","created_at":"2026-05-20T00:04:58Z"}],"graph_snapshots":[{"event_id":"sha256:b8c9fd3b7f2bfd1b18f4918442c4665576a39bf6002a0e6ff8844fa343f42656","target":"graph","created_at":"2026-05-20T00:04:58Z","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/2605.17787/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"It is widely believed that stochastic gradient descent (SGD) performs significantly worse than adaptive optimizers such as Adam in pre-training Large Language Models (LLMs). Yet the underlying reason for this gap remains unclear. In this work, we attribute a large part of the discrepancy to SGD's inability to sustain learning rates comparable to Adam's much larger effective learning rates. Through empirical and theoretical analysis of LLM pre-training dynamics, we identify that training is characterized by small gradient norms and large weight-to-gradient ratios, an effect that becomes more pr","authors_text":"Athanasios Glentis, Chung-Yiu Yau, Dawei Li, Mingyi Hong","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T03:09:50Z","title":"Revisiting the Adam-SGD Gap in LLM Pre-Training: The Role of Large Effective Learning Rates"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.17787","kind":"arxiv","version":1},"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":"sha256:1ea079e4d5c0e898d1b5eca7cf2a156dad3eeb09fdb96065cd05330667d94198","target":"record","created_at":"2026-05-20T00:04:58Z","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":"7b316213f566d638a7753779aa2379a393ac9a132b04e258c0fafc07ec7cb652","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T03:09:50Z","title_canon_sha256":"358fb531365be932999f0ea161178ce250f8ee0cbb25b031c7930c433a14d61f"},"schema_version":"1.0","source":{"id":"2605.17787","kind":"arxiv","version":1}},"canonical_sha256":"828b82ddc3cd37628c12a4a6a0357c8615bf27df981132dc0527ccb410c8b651","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"828b82ddc3cd37628c12a4a6a0357c8615bf27df981132dc0527ccb410c8b651","first_computed_at":"2026-05-20T00:04:58.313805Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:04:58.313805Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"CGtmcl1wcbf2mMwedEQYsdqjf+/IZPBZDNifhb17W1c33gIc3aYvBdY591SMV3hg0c1QjNJVgnmd2D01RonaAA==","signature_status":"signed_v1","signed_at":"2026-05-20T00:04:58.314616Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.17787","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1ea079e4d5c0e898d1b5eca7cf2a156dad3eeb09fdb96065cd05330667d94198","sha256:b8c9fd3b7f2bfd1b18f4918442c4665576a39bf6002a0e6ff8844fa343f42656"],"state_sha256":"be8bbaa97693a09eeecf81f2a182f3695579cee605a9bcc4a296abfeeb0a33f8"}