{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ECXVOYFLWFEG5VNRRDTUMVYYRT","short_pith_number":"pith:ECXVOYFL","schema_version":"1.0","canonical_sha256":"20af5760abb1486ed5b188e74657188cc2b6ebacd0a1cefebdaf16108079d9c3","source":{"kind":"arxiv","id":"2605.20798","version":1},"attestation_state":"computed","paper":{"title":"Most Transformer Modifications Still Do Not Transfer at 1-3B: A 2020-2026 Update to Narang et al. (2021) with Downstream Evaluation and a Noise Floor","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Bin Huang, Guhua Zhang, Jiahao Lu, Jie Zhou, Yang Zhao","submitted_at":"2026-05-20T06:43:34Z","abstract_excerpt":"Narang et al. (2021) evaluated 40+ Transformer modifications at T5-base scale and concluded that most did not transfer. Five years later, the typical working regime has moved to 1-3B parameters, downstream evaluation has replaced pretraining perplexity, and a substantially different catalogue of modifications has emerged. We revisit their question by testing 20 post-2021 Transformer modifications at 1.2B and 3B under strict iso-data, iso-compute, iso-recipe control, with a multi-seed baseline noise floor and CLIMB-12 downstream evaluation as the primary metric. The central finding reproduces t"},"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":"2605.20798","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T06:43:34Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"3e397e40d955eaee4b866651f38dcbd12df1fea433f25c629ca48aebbb2548d7","abstract_canon_sha256":"74fae8c1a7c6151f130379309784ef8459b3d91153f3dce90dacd627619ba062"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:04:55.038199Z","signature_b64":"sM17YGJl/eZ1okeGxC2vzNjmZdm+kXL2SCo0FBJQAmchdEUJU3E6d0uazP2WesPMJaTO5GRnMPF+VL8mRJsrAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"20af5760abb1486ed5b188e74657188cc2b6ebacd0a1cefebdaf16108079d9c3","last_reissued_at":"2026-05-21T01:04:55.037425Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:04:55.037425Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Most Transformer Modifications Still Do Not Transfer at 1-3B: A 2020-2026 Update to Narang et al. (2021) with Downstream Evaluation and a Noise Floor","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Bin Huang, Guhua Zhang, Jiahao Lu, Jie Zhou, Yang Zhao","submitted_at":"2026-05-20T06:43:34Z","abstract_excerpt":"Narang et al. (2021) evaluated 40+ Transformer modifications at T5-base scale and concluded that most did not transfer. Five years later, the typical working regime has moved to 1-3B parameters, downstream evaluation has replaced pretraining perplexity, and a substantially different catalogue of modifications has emerged. We revisit their question by testing 20 post-2021 Transformer modifications at 1.2B and 3B under strict iso-data, iso-compute, iso-recipe control, with a multi-seed baseline noise floor and CLIMB-12 downstream evaluation as the primary metric. The central finding reproduces t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20798","kind":"arxiv","version":1},"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/2605.20798/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":"2605.20798","created_at":"2026-05-21T01:04:55.037552+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.20798v1","created_at":"2026-05-21T01:04:55.037552+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20798","created_at":"2026-05-21T01:04:55.037552+00:00"},{"alias_kind":"pith_short_12","alias_value":"ECXVOYFLWFEG","created_at":"2026-05-21T01:04:55.037552+00:00"},{"alias_kind":"pith_short_16","alias_value":"ECXVOYFLWFEG5VNR","created_at":"2026-05-21T01:04:55.037552+00:00"},{"alias_kind":"pith_short_8","alias_value":"ECXVOYFL","created_at":"2026-05-21T01:04:55.037552+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ECXVOYFLWFEG5VNRRDTUMVYYRT","json":"https://pith.science/pith/ECXVOYFLWFEG5VNRRDTUMVYYRT.json","graph_json":"https://pith.science/api/pith-number/ECXVOYFLWFEG5VNRRDTUMVYYRT/graph.json","events_json":"https://pith.science/api/pith-number/ECXVOYFLWFEG5VNRRDTUMVYYRT/events.json","paper":"https://pith.science/paper/ECXVOYFL"},"agent_actions":{"view_html":"https://pith.science/pith/ECXVOYFLWFEG5VNRRDTUMVYYRT","download_json":"https://pith.science/pith/ECXVOYFLWFEG5VNRRDTUMVYYRT.json","view_paper":"https://pith.science/paper/ECXVOYFL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.20798&json=true","fetch_graph":"https://pith.science/api/pith-number/ECXVOYFLWFEG5VNRRDTUMVYYRT/graph.json","fetch_events":"https://pith.science/api/pith-number/ECXVOYFLWFEG5VNRRDTUMVYYRT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ECXVOYFLWFEG5VNRRDTUMVYYRT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ECXVOYFLWFEG5VNRRDTUMVYYRT/action/storage_attestation","attest_author":"https://pith.science/pith/ECXVOYFLWFEG5VNRRDTUMVYYRT/action/author_attestation","sign_citation":"https://pith.science/pith/ECXVOYFLWFEG5VNRRDTUMVYYRT/action/citation_signature","submit_replication":"https://pith.science/pith/ECXVOYFLWFEG5VNRRDTUMVYYRT/action/replication_record"}},"created_at":"2026-05-21T01:04:55.037552+00:00","updated_at":"2026-05-21T01:04:55.037552+00:00"}