{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3PXYRYNZ4HCSFDJATADPKHA24A","short_pith_number":"pith:3PXYRYNZ","schema_version":"1.0","canonical_sha256":"dbef88e1b9e1c5228d209806f51c1ae01c43f92ab60a234cd67de78c49a829e8","source":{"kind":"arxiv","id":"2408.04270","version":1},"attestation_state":"computed","paper":{"title":"Analysis of Argument Structure Constructions in the Large Language Model BERT","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Achim Schilling, Patrick Krauss, Pegah Ramezani","submitted_at":"2024-08-08T07:12:46Z","abstract_excerpt":"This study investigates how BERT processes and represents Argument Structure Constructions (ASCs), extending previous LSTM analyses. Using a dataset of 2000 sentences across four ASC types (transitive, ditransitive, caused-motion, resultative), we analyzed BERT's token embeddings across 12 layers. Visualizations with MDS and t-SNE and clustering quantified by Generalized Discrimination Value (GDV) were used. Feedforward classifiers (probes) predicted construction categories from embeddings. CLS token embeddings clustered best in layers 2-4, decreased in intermediate layers, and slightly increa"},"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":"2408.04270","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-08-08T07:12:46Z","cross_cats_sorted":[],"title_canon_sha256":"f8e74047b797606083dbbf62be0a1c6fba2ffe499f1f04ff0dc4460936538db9","abstract_canon_sha256":"3c6779df07da0c3b2987475cd31e0f95c8fc6019d2893ef50f15077a72a0697b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:53:30.792981Z","signature_b64":"4qgDUqqeIaliTPYrEifzQIArp+FYtOhOXhiG84Zu7q9q/2+Hi8L8R0nzutkT/EyoI1b06BbCfjM8yT6ccWFTAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dbef88e1b9e1c5228d209806f51c1ae01c43f92ab60a234cd67de78c49a829e8","last_reissued_at":"2026-07-05T08:53:30.792507Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:53:30.792507Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Analysis of Argument Structure Constructions in the Large Language Model BERT","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Achim Schilling, Patrick Krauss, Pegah Ramezani","submitted_at":"2024-08-08T07:12:46Z","abstract_excerpt":"This study investigates how BERT processes and represents Argument Structure Constructions (ASCs), extending previous LSTM analyses. Using a dataset of 2000 sentences across four ASC types (transitive, ditransitive, caused-motion, resultative), we analyzed BERT's token embeddings across 12 layers. Visualizations with MDS and t-SNE and clustering quantified by Generalized Discrimination Value (GDV) were used. Feedforward classifiers (probes) predicted construction categories from embeddings. CLS token embeddings clustered best in layers 2-4, decreased in intermediate layers, and slightly increa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.04270","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/2408.04270/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":"2408.04270","created_at":"2026-07-05T08:53:30.792563+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.04270v1","created_at":"2026-07-05T08:53:30.792563+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.04270","created_at":"2026-07-05T08:53:30.792563+00:00"},{"alias_kind":"pith_short_12","alias_value":"3PXYRYNZ4HCS","created_at":"2026-07-05T08:53:30.792563+00:00"},{"alias_kind":"pith_short_16","alias_value":"3PXYRYNZ4HCSFDJA","created_at":"2026-07-05T08:53:30.792563+00:00"},{"alias_kind":"pith_short_8","alias_value":"3PXYRYNZ","created_at":"2026-07-05T08:53:30.792563+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/3PXYRYNZ4HCSFDJATADPKHA24A","json":"https://pith.science/pith/3PXYRYNZ4HCSFDJATADPKHA24A.json","graph_json":"https://pith.science/api/pith-number/3PXYRYNZ4HCSFDJATADPKHA24A/graph.json","events_json":"https://pith.science/api/pith-number/3PXYRYNZ4HCSFDJATADPKHA24A/events.json","paper":"https://pith.science/paper/3PXYRYNZ"},"agent_actions":{"view_html":"https://pith.science/pith/3PXYRYNZ4HCSFDJATADPKHA24A","download_json":"https://pith.science/pith/3PXYRYNZ4HCSFDJATADPKHA24A.json","view_paper":"https://pith.science/paper/3PXYRYNZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.04270&json=true","fetch_graph":"https://pith.science/api/pith-number/3PXYRYNZ4HCSFDJATADPKHA24A/graph.json","fetch_events":"https://pith.science/api/pith-number/3PXYRYNZ4HCSFDJATADPKHA24A/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3PXYRYNZ4HCSFDJATADPKHA24A/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3PXYRYNZ4HCSFDJATADPKHA24A/action/storage_attestation","attest_author":"https://pith.science/pith/3PXYRYNZ4HCSFDJATADPKHA24A/action/author_attestation","sign_citation":"https://pith.science/pith/3PXYRYNZ4HCSFDJATADPKHA24A/action/citation_signature","submit_replication":"https://pith.science/pith/3PXYRYNZ4HCSFDJATADPKHA24A/action/replication_record"}},"created_at":"2026-07-05T08:53:30.792563+00:00","updated_at":"2026-07-05T08:53:30.792563+00:00"}