{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:76BGFZKK5NALYB5IDX2333BACX","short_pith_number":"pith:76BGFZKK","schema_version":"1.0","canonical_sha256":"ff8262e54aeb40bc07a81df5bdec2015ffb0eb237f3db1e7025f1569cdeaac85","source":{"kind":"arxiv","id":"2601.09719","version":3},"attestation_state":"computed","paper":{"title":"Bounded Hyperbolic Tangent: A Stable and Efficient Alternative to Pre-Layer Normalization in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hoyoon Byun, Kyungwoo Song, Sungrae Park, Taero Kim, Youngjun Choi","submitted_at":"2025-12-26T06:22:13Z","abstract_excerpt":"Pre-Layer Normalization (Pre-LN) is the de facto choice for large language models (LLMs) and is crucial for stable pretraining and effective transfer learning. However, Pre-LN incurs repeated statistical-computation overhead and remains vulnerable to the curse of depth, where hidden-state magnitudes and variances grow as the number of layers increases, destabilizing training. Efficiency-oriented normalization-free methods such as Dynamic Tanh (DyT) improve throughput but remain fragile at depth. To jointly address stability and efficiency, we propose Bounded Hyperbolic Tanh (BHyT), a drop-in r"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2601.09719","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-12-26T06:22:13Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"df7abdf6d515e2a6bdb2d584a5001bd00bc4304fe9653ea3549076733f7327ae","abstract_canon_sha256":"d742b22a8d3a19096c37bdd0cba4cd9dc11ee5d9b7a7a21db1f8271c796ef3d7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-04T01:09:41.044351Z","signature_b64":"5yK2iciWAwOxlsHEM/YLAk0N/4/hpDigg8qEu8oJz3JiNL8+ZQJBaQISgpYcz/Yu4He/IHp3EaXJKtaH24CcCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ff8262e54aeb40bc07a81df5bdec2015ffb0eb237f3db1e7025f1569cdeaac85","last_reissued_at":"2026-06-04T01:09:41.043792Z","signature_status":"signed_v1","first_computed_at":"2026-06-04T01:09:41.043792Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bounded Hyperbolic Tangent: A Stable and Efficient Alternative to Pre-Layer Normalization in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hoyoon Byun, Kyungwoo Song, Sungrae Park, Taero Kim, Youngjun Choi","submitted_at":"2025-12-26T06:22:13Z","abstract_excerpt":"Pre-Layer Normalization (Pre-LN) is the de facto choice for large language models (LLMs) and is crucial for stable pretraining and effective transfer learning. However, Pre-LN incurs repeated statistical-computation overhead and remains vulnerable to the curse of depth, where hidden-state magnitudes and variances grow as the number of layers increases, destabilizing training. Efficiency-oriented normalization-free methods such as Dynamic Tanh (DyT) improve throughput but remain fragile at depth. To jointly address stability and efficiency, we propose Bounded Hyperbolic Tanh (BHyT), a drop-in r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.09719","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/2601.09719/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2601.09719","created_at":"2026-06-04T01:09:41.043857+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.09719v3","created_at":"2026-06-04T01:09:41.043857+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.09719","created_at":"2026-06-04T01:09:41.043857+00:00"},{"alias_kind":"pith_short_12","alias_value":"76BGFZKK5NAL","created_at":"2026-06-04T01:09:41.043857+00:00"},{"alias_kind":"pith_short_16","alias_value":"76BGFZKK5NALYB5I","created_at":"2026-06-04T01:09:41.043857+00:00"},{"alias_kind":"pith_short_8","alias_value":"76BGFZKK","created_at":"2026-06-04T01:09:41.043857+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.23434","citing_title":"When Does Removing LayerNorm Help? Activation Bounding as a Regime-Dependent Implicit Regularizer","ref_index":4,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/76BGFZKK5NALYB5IDX2333BACX","json":"https://pith.science/pith/76BGFZKK5NALYB5IDX2333BACX.json","graph_json":"https://pith.science/api/pith-number/76BGFZKK5NALYB5IDX2333BACX/graph.json","events_json":"https://pith.science/api/pith-number/76BGFZKK5NALYB5IDX2333BACX/events.json","paper":"https://pith.science/paper/76BGFZKK"},"agent_actions":{"view_html":"https://pith.science/pith/76BGFZKK5NALYB5IDX2333BACX","download_json":"https://pith.science/pith/76BGFZKK5NALYB5IDX2333BACX.json","view_paper":"https://pith.science/paper/76BGFZKK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.09719&json=true","fetch_graph":"https://pith.science/api/pith-number/76BGFZKK5NALYB5IDX2333BACX/graph.json","fetch_events":"https://pith.science/api/pith-number/76BGFZKK5NALYB5IDX2333BACX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/76BGFZKK5NALYB5IDX2333BACX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/76BGFZKK5NALYB5IDX2333BACX/action/storage_attestation","attest_author":"https://pith.science/pith/76BGFZKK5NALYB5IDX2333BACX/action/author_attestation","sign_citation":"https://pith.science/pith/76BGFZKK5NALYB5IDX2333BACX/action/citation_signature","submit_replication":"https://pith.science/pith/76BGFZKK5NALYB5IDX2333BACX/action/replication_record"}},"created_at":"2026-06-04T01:09:41.043857+00:00","updated_at":"2026-06-04T01:09:41.043857+00:00"}