{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ALKCTMJHYGUKXWCJCG5CX6BBAR","short_pith_number":"pith:ALKCTMJH","schema_version":"1.0","canonical_sha256":"02d429b127c1a8abd84911ba2bf82104777a5b51d9467997ec54db15228ad671","source":{"kind":"arxiv","id":"2602.10408","version":2},"attestation_state":"computed","paper":{"title":"Gated Normalization Removal and Scale Anchoring in Pre-Norm Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Andrei Kanavalau, Carmen Amo Alonso, Sanjay Lall","submitted_at":"2026-02-11T01:40:34Z","abstract_excerpt":"Normalization layers are standard in transformers, but it is not clear whether their sample-dependent computations are necessary throughout both training and inference. This work develops a gated normalization-removal approach for pre-norm transformers. The approach is implemented using TaperNorm, which starts from standard RMSNorm/LayerNorm and gradually tapers to learned sample-independent linear or affine maps. Once the gate reaches zero, per-token statistics are no longer computed in the tapered layers and the resulting maps can be folded into adjacent linear projections. The results indic"},"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":"2602.10408","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-11T01:40:34Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"ce0b5d760adad3732b63c49c2cc82adfa9971924bbd21d9d1aea950735990fad","abstract_canon_sha256":"e6b894b0e653cbb1a1150c4a4ed2465f7ee3376e3485afe3018b791f580591d0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:04:22.897625Z","signature_b64":"pC2L4zJqG4d6fna05Zbnm3bb1DEQuUDGC5ifCg+3mhzX1ZsWIcg5rakTcjFHU4H8acegi5dUgCBtuIBvmttPAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"02d429b127c1a8abd84911ba2bf82104777a5b51d9467997ec54db15228ad671","last_reissued_at":"2026-05-21T01:04:22.896796Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:04:22.896796Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gated Normalization Removal and Scale Anchoring in Pre-Norm Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Andrei Kanavalau, Carmen Amo Alonso, Sanjay Lall","submitted_at":"2026-02-11T01:40:34Z","abstract_excerpt":"Normalization layers are standard in transformers, but it is not clear whether their sample-dependent computations are necessary throughout both training and inference. This work develops a gated normalization-removal approach for pre-norm transformers. The approach is implemented using TaperNorm, which starts from standard RMSNorm/LayerNorm and gradually tapers to learned sample-independent linear or affine maps. Once the gate reaches zero, per-token statistics are no longer computed in the tapered layers and the resulting maps can be folded into adjacent linear projections. The results indic"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.10408","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/2602.10408/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":"2602.10408","created_at":"2026-05-21T01:04:22.896890+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.10408v2","created_at":"2026-05-21T01:04:22.896890+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.10408","created_at":"2026-05-21T01:04:22.896890+00:00"},{"alias_kind":"pith_short_12","alias_value":"ALKCTMJHYGUK","created_at":"2026-05-21T01:04:22.896890+00:00"},{"alias_kind":"pith_short_16","alias_value":"ALKCTMJHYGUKXWCJ","created_at":"2026-05-21T01:04:22.896890+00:00"},{"alias_kind":"pith_short_8","alias_value":"ALKCTMJH","created_at":"2026-05-21T01:04:22.896890+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":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ALKCTMJHYGUKXWCJCG5CX6BBAR","json":"https://pith.science/pith/ALKCTMJHYGUKXWCJCG5CX6BBAR.json","graph_json":"https://pith.science/api/pith-number/ALKCTMJHYGUKXWCJCG5CX6BBAR/graph.json","events_json":"https://pith.science/api/pith-number/ALKCTMJHYGUKXWCJCG5CX6BBAR/events.json","paper":"https://pith.science/paper/ALKCTMJH"},"agent_actions":{"view_html":"https://pith.science/pith/ALKCTMJHYGUKXWCJCG5CX6BBAR","download_json":"https://pith.science/pith/ALKCTMJHYGUKXWCJCG5CX6BBAR.json","view_paper":"https://pith.science/paper/ALKCTMJH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.10408&json=true","fetch_graph":"https://pith.science/api/pith-number/ALKCTMJHYGUKXWCJCG5CX6BBAR/graph.json","fetch_events":"https://pith.science/api/pith-number/ALKCTMJHYGUKXWCJCG5CX6BBAR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ALKCTMJHYGUKXWCJCG5CX6BBAR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ALKCTMJHYGUKXWCJCG5CX6BBAR/action/storage_attestation","attest_author":"https://pith.science/pith/ALKCTMJHYGUKXWCJCG5CX6BBAR/action/author_attestation","sign_citation":"https://pith.science/pith/ALKCTMJHYGUKXWCJCG5CX6BBAR/action/citation_signature","submit_replication":"https://pith.science/pith/ALKCTMJHYGUKXWCJCG5CX6BBAR/action/replication_record"}},"created_at":"2026-05-21T01:04:22.896890+00:00","updated_at":"2026-05-21T01:04:22.896890+00:00"}