{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:DGN7Q33F3Q3G6J33FNDSIFQSDM","short_pith_number":"pith:DGN7Q33F","canonical_record":{"source":{"id":"2502.15499","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-21T14:49:34Z","cross_cats_sorted":[],"title_canon_sha256":"fb3acf45b4df9339c07f7e15c83d31051ae5b50350826badbcb666717fc8e188","abstract_canon_sha256":"5a85d82398ebb5c0a26707cf83430f797fcd00c70a0bd7b4b9b3bc962bb80f92"},"schema_version":"1.0"},"canonical_sha256":"199bf86f65dc366f277b2b472416121b2c13dc6bb05fcc93d9f7752379ffc242","source":{"kind":"arxiv","id":"2502.15499","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.15499","created_at":"2026-07-05T10:19:39Z"},{"alias_kind":"arxiv_version","alias_value":"2502.15499v2","created_at":"2026-07-05T10:19:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.15499","created_at":"2026-07-05T10:19:39Z"},{"alias_kind":"pith_short_12","alias_value":"DGN7Q33F3Q3G","created_at":"2026-07-05T10:19:39Z"},{"alias_kind":"pith_short_16","alias_value":"DGN7Q33F3Q3G6J33","created_at":"2026-07-05T10:19:39Z"},{"alias_kind":"pith_short_8","alias_value":"DGN7Q33F","created_at":"2026-07-05T10:19:39Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:DGN7Q33F3Q3G6J33FNDSIFQSDM","target":"record","payload":{"canonical_record":{"source":{"id":"2502.15499","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-21T14:49:34Z","cross_cats_sorted":[],"title_canon_sha256":"fb3acf45b4df9339c07f7e15c83d31051ae5b50350826badbcb666717fc8e188","abstract_canon_sha256":"5a85d82398ebb5c0a26707cf83430f797fcd00c70a0bd7b4b9b3bc962bb80f92"},"schema_version":"1.0"},"canonical_sha256":"199bf86f65dc366f277b2b472416121b2c13dc6bb05fcc93d9f7752379ffc242","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:19:39.830206Z","signature_b64":"Ds/L7IGkgVAx3EEHUacwVFuyqTWlr1OipoiCjevWYQ//m2cRDwECJwRAZul04nsUFR69/Up/jmKTo121SvvUDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"199bf86f65dc366f277b2b472416121b2c13dc6bb05fcc93d9f7752379ffc242","last_reissued_at":"2026-07-05T10:19:39.829657Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:19:39.829657Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.15499","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":"2026-07-05T10:19:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"AYVUya6Xa5VEfsTWvUwkrY6O9zFGU9+I7qSiytl0CX2HF7dPS5bEHr7zJ35jkcCtUNeRiXfVbhyGwUacZ39TDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T19:14:12.362309Z"},"content_sha256":"9182936d933c627925051dba5dd0d7a21ee8866b262dc1d319834fa0a3a893b4","schema_version":"1.0","event_id":"sha256:9182936d933c627925051dba5dd0d7a21ee8866b262dc1d319834fa0a3a893b4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:DGN7Q33F3Q3G6J33FNDSIFQSDM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Scale-Distribution Decoupling: Enabling Stable and Effective Training of Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jian Yang, Xiaoqing Li, Xun Zhou, Ya Wang, Yutao Zeng, Zhijian Zhuo","submitted_at":"2025-02-21T14:49:34Z","abstract_excerpt":"Training stability is a persistent challenge in the pre-training of large language models (LLMs), particularly for architectures such as Post-Norm Transformers, which are prone to gradient explosion and dissipation. In this paper, we propose Scale-Distribution Decoupling (SDD), a novel approach that stabilizes training by explicitly decoupling the scale and distribution of the weight matrix in fully-connected layers. SDD applies a normalization mechanism to regulate activations and a learnable scaling vector to maintain well-conditioned gradients, effectively preventing $\\textbf{gradient explo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.15499","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/2502.15499/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":"2026-07-05T10:19:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"30UlnzcZDeEtnTIHIO6pe9QSM+0U/KqYQDaIeIs4+svKBWcAL2IG/DIZs3VpLHVtcDMZRsdcOZKbJrwfx5OVDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T19:14:12.362802Z"},"content_sha256":"ca69a69ab643914cc7ba848b27307d391fdba0b5110c23d3f7026c8a3c4beb06","schema_version":"1.0","event_id":"sha256:ca69a69ab643914cc7ba848b27307d391fdba0b5110c23d3f7026c8a3c4beb06"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DGN7Q33F3Q3G6J33FNDSIFQSDM/bundle.json","state_url":"https://pith.science/pith/DGN7Q33F3Q3G6J33FNDSIFQSDM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DGN7Q33F3Q3G6J33FNDSIFQSDM/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":"2026-08-03T19:14:12Z","links":{"resolver":"https://pith.science/pith/DGN7Q33F3Q3G6J33FNDSIFQSDM","bundle":"https://pith.science/pith/DGN7Q33F3Q3G6J33FNDSIFQSDM/bundle.json","state":"https://pith.science/pith/DGN7Q33F3Q3G6J33FNDSIFQSDM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DGN7Q33F3Q3G6J33FNDSIFQSDM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:DGN7Q33F3Q3G6J33FNDSIFQSDM","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":"5a85d82398ebb5c0a26707cf83430f797fcd00c70a0bd7b4b9b3bc962bb80f92","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-21T14:49:34Z","title_canon_sha256":"fb3acf45b4df9339c07f7e15c83d31051ae5b50350826badbcb666717fc8e188"},"schema_version":"1.0","source":{"id":"2502.15499","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.15499","created_at":"2026-07-05T10:19:39Z"},{"alias_kind":"arxiv_version","alias_value":"2502.15499v2","created_at":"2026-07-05T10:19:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.15499","created_at":"2026-07-05T10:19:39Z"},{"alias_kind":"pith_short_12","alias_value":"DGN7Q33F3Q3G","created_at":"2026-07-05T10:19:39Z"},{"alias_kind":"pith_short_16","alias_value":"DGN7Q33F3Q3G6J33","created_at":"2026-07-05T10:19:39Z"},{"alias_kind":"pith_short_8","alias_value":"DGN7Q33F","created_at":"2026-07-05T10:19:39Z"}],"graph_snapshots":[{"event_id":"sha256:ca69a69ab643914cc7ba848b27307d391fdba0b5110c23d3f7026c8a3c4beb06","target":"graph","created_at":"2026-07-05T10:19:39Z","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/2502.15499/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Training stability is a persistent challenge in the pre-training of large language models (LLMs), particularly for architectures such as Post-Norm Transformers, which are prone to gradient explosion and dissipation. In this paper, we propose Scale-Distribution Decoupling (SDD), a novel approach that stabilizes training by explicitly decoupling the scale and distribution of the weight matrix in fully-connected layers. SDD applies a normalization mechanism to regulate activations and a learnable scaling vector to maintain well-conditioned gradients, effectively preventing $\\textbf{gradient explo","authors_text":"Jian Yang, Xiaoqing Li, Xun Zhou, Ya Wang, Yutao Zeng, Zhijian Zhuo","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-21T14:49:34Z","title":"Scale-Distribution Decoupling: Enabling Stable and Effective Training of Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.15499","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":"sha256:9182936d933c627925051dba5dd0d7a21ee8866b262dc1d319834fa0a3a893b4","target":"record","created_at":"2026-07-05T10:19:39Z","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":"5a85d82398ebb5c0a26707cf83430f797fcd00c70a0bd7b4b9b3bc962bb80f92","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-21T14:49:34Z","title_canon_sha256":"fb3acf45b4df9339c07f7e15c83d31051ae5b50350826badbcb666717fc8e188"},"schema_version":"1.0","source":{"id":"2502.15499","kind":"arxiv","version":2}},"canonical_sha256":"199bf86f65dc366f277b2b472416121b2c13dc6bb05fcc93d9f7752379ffc242","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"199bf86f65dc366f277b2b472416121b2c13dc6bb05fcc93d9f7752379ffc242","first_computed_at":"2026-07-05T10:19:39.829657Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:19:39.829657Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Ds/L7IGkgVAx3EEHUacwVFuyqTWlr1OipoiCjevWYQ//m2cRDwECJwRAZul04nsUFR69/Up/jmKTo121SvvUDA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:19:39.830206Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.15499","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9182936d933c627925051dba5dd0d7a21ee8866b262dc1d319834fa0a3a893b4","sha256:ca69a69ab643914cc7ba848b27307d391fdba0b5110c23d3f7026c8a3c4beb06"],"state_sha256":"d0b6e232767324b7590cac3d8d4d91d10d7b684d4c3018d69b29d79945666cb2"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WOSQx2ijCaQ2EwA+R/V23u6CSf6nXJAlOf+SLiEukRrr2pr94Q1LtfhPxZtowwqtETTrccqAZrYNZuMbBsulDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-03T19:14:12.366204Z","bundle_sha256":"0f316e1e19115fbb433cc4fd3d748524b5d633b5cc74fa24cb14c11ee582bc91"}}