{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:J2OQ3OVKWCMXLXF3TNHBAQADPP","short_pith_number":"pith:J2OQ3OVK","schema_version":"1.0","canonical_sha256":"4e9d0dbaaab09975dcbb9b4e1040037bdebb0aa50e59acffd0dc143e6d060979","source":{"kind":"arxiv","id":"2403.02436","version":3},"attestation_state":"computed","paper":{"title":"How does Architecture Influence the Base Capabilities of Pre-trained Language Models? A Case Study Based on FFN-Wider and MoE Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bing Qin, Dongliang Xu, Liangyu Huo, Qing Yang, Xin Lu, Yanyan Zhao","submitted_at":"2024-03-04T19:33:39Z","abstract_excerpt":"Pre-trained language models have been proven to possess strong base capabilities, which not only excel in in-distribution language modeling but also show powerful abilities in out-of-distribution language modeling, transfer learning and few-shot learning. Unlike existing work focusing on the influence of scale on base capabilities, our work examines the influence of architecture on those. Specifically, our concern is: How does architecture influence the base capabilities of pre-trained language models? In this work, we attempt to explain and reverse the decline in base capabilities caused by 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":"2403.02436","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-04T19:33:39Z","cross_cats_sorted":[],"title_canon_sha256":"e64b0d628c48f11ebf5a06852f0a77bbdb2b8c355d1abbe96d50b849a86cb931","abstract_canon_sha256":"1ec5d992152c99db5faf5c2bb061373c89cc899fb645e27254d7d5a9c455bd8b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:28:48.395455Z","signature_b64":"mrCb3f/6G++vi2hAgF1Gq1kgFwFUdGSQt1ynnK5rBcxJMCy7vpXT+3pryOfXYUO6fIR6cADxvOE05Eqmw6g6AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4e9d0dbaaab09975dcbb9b4e1040037bdebb0aa50e59acffd0dc143e6d060979","last_reissued_at":"2026-07-05T09:28:48.394960Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:28:48.394960Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How does Architecture Influence the Base Capabilities of Pre-trained Language Models? A Case Study Based on FFN-Wider and MoE Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bing Qin, Dongliang Xu, Liangyu Huo, Qing Yang, Xin Lu, Yanyan Zhao","submitted_at":"2024-03-04T19:33:39Z","abstract_excerpt":"Pre-trained language models have been proven to possess strong base capabilities, which not only excel in in-distribution language modeling but also show powerful abilities in out-of-distribution language modeling, transfer learning and few-shot learning. Unlike existing work focusing on the influence of scale on base capabilities, our work examines the influence of architecture on those. Specifically, our concern is: How does architecture influence the base capabilities of pre-trained language models? In this work, we attempt to explain and reverse the decline in base capabilities caused by t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.02436","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/2403.02436/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":"2403.02436","created_at":"2026-07-05T09:28:48.395019+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.02436v3","created_at":"2026-07-05T09:28:48.395019+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.02436","created_at":"2026-07-05T09:28:48.395019+00:00"},{"alias_kind":"pith_short_12","alias_value":"J2OQ3OVKWCMX","created_at":"2026-07-05T09:28:48.395019+00:00"},{"alias_kind":"pith_short_16","alias_value":"J2OQ3OVKWCMXLXF3","created_at":"2026-07-05T09:28:48.395019+00:00"},{"alias_kind":"pith_short_8","alias_value":"J2OQ3OVK","created_at":"2026-07-05T09:28:48.395019+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/J2OQ3OVKWCMXLXF3TNHBAQADPP","json":"https://pith.science/pith/J2OQ3OVKWCMXLXF3TNHBAQADPP.json","graph_json":"https://pith.science/api/pith-number/J2OQ3OVKWCMXLXF3TNHBAQADPP/graph.json","events_json":"https://pith.science/api/pith-number/J2OQ3OVKWCMXLXF3TNHBAQADPP/events.json","paper":"https://pith.science/paper/J2OQ3OVK"},"agent_actions":{"view_html":"https://pith.science/pith/J2OQ3OVKWCMXLXF3TNHBAQADPP","download_json":"https://pith.science/pith/J2OQ3OVKWCMXLXF3TNHBAQADPP.json","view_paper":"https://pith.science/paper/J2OQ3OVK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.02436&json=true","fetch_graph":"https://pith.science/api/pith-number/J2OQ3OVKWCMXLXF3TNHBAQADPP/graph.json","fetch_events":"https://pith.science/api/pith-number/J2OQ3OVKWCMXLXF3TNHBAQADPP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/J2OQ3OVKWCMXLXF3TNHBAQADPP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/J2OQ3OVKWCMXLXF3TNHBAQADPP/action/storage_attestation","attest_author":"https://pith.science/pith/J2OQ3OVKWCMXLXF3TNHBAQADPP/action/author_attestation","sign_citation":"https://pith.science/pith/J2OQ3OVKWCMXLXF3TNHBAQADPP/action/citation_signature","submit_replication":"https://pith.science/pith/J2OQ3OVKWCMXLXF3TNHBAQADPP/action/replication_record"}},"created_at":"2026-07-05T09:28:48.395019+00:00","updated_at":"2026-07-05T09:28:48.395019+00:00"}