{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3QAMPF5KPRPJBTRBLHJNYHLGI3","short_pith_number":"pith:3QAMPF5K","schema_version":"1.0","canonical_sha256":"dc00c797aa7c5e90ce2159d2dc1d6646efa448c9eaadea7b783b3d6f80bafbb5","source":{"kind":"arxiv","id":"2403.15938","version":1},"attestation_state":"computed","paper":{"title":"LlamBERT: Large-scale low-cost data annotation in NLP","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Andr\\'as Luk\\'acs, B\\'alint Csan\\'ady, Lajos Muzsai, P\\'eter Vedres, Zolt\\'an N\\'adasdy","submitted_at":"2024-03-23T21:54:34Z","abstract_excerpt":"Large Language Models (LLMs), such as GPT-4 and Llama 2, show remarkable proficiency in a wide range of natural language processing (NLP) tasks. Despite their effectiveness, the high costs associated with their use pose a challenge. We present LlamBERT, a hybrid approach that leverages LLMs to annotate a small subset of large, unlabeled databases and uses the results for fine-tuning transformer encoders like BERT and RoBERTa. This strategy is evaluated on two diverse datasets: the IMDb review dataset and the UMLS Meta-Thesaurus. Our results indicate that the LlamBERT approach slightly compromi"},"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":"2403.15938","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-23T21:54:34Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"0aedeeead6ee8ce2b1133bc55a43a33ab79ab9ea7f0a1b2f1034f7dfb50de6a9","abstract_canon_sha256":"aca7860a84c7f5393ea095749196c5f62e743c5676bb6ec7bb405406e6378d8f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:00:08.650452Z","signature_b64":"kp36DgnNIFh+4+CN6qLgkOfJinVAxrvR6IyowAWplDBRGx2DWmB5nsKCqb5w75evtx4BovsR5bIAhhYMSJHcDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dc00c797aa7c5e90ce2159d2dc1d6646efa448c9eaadea7b783b3d6f80bafbb5","last_reissued_at":"2026-07-05T08:00:08.649914Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:00:08.649914Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LlamBERT: Large-scale low-cost data annotation in NLP","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Andr\\'as Luk\\'acs, B\\'alint Csan\\'ady, Lajos Muzsai, P\\'eter Vedres, Zolt\\'an N\\'adasdy","submitted_at":"2024-03-23T21:54:34Z","abstract_excerpt":"Large Language Models (LLMs), such as GPT-4 and Llama 2, show remarkable proficiency in a wide range of natural language processing (NLP) tasks. Despite their effectiveness, the high costs associated with their use pose a challenge. We present LlamBERT, a hybrid approach that leverages LLMs to annotate a small subset of large, unlabeled databases and uses the results for fine-tuning transformer encoders like BERT and RoBERTa. This strategy is evaluated on two diverse datasets: the IMDb review dataset and the UMLS Meta-Thesaurus. Our results indicate that the LlamBERT approach slightly compromi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.15938","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/2403.15938/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":"2403.15938","created_at":"2026-07-05T08:00:08.650000+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.15938v1","created_at":"2026-07-05T08:00:08.650000+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.15938","created_at":"2026-07-05T08:00:08.650000+00:00"},{"alias_kind":"pith_short_12","alias_value":"3QAMPF5KPRPJ","created_at":"2026-07-05T08:00:08.650000+00:00"},{"alias_kind":"pith_short_16","alias_value":"3QAMPF5KPRPJBTRB","created_at":"2026-07-05T08:00:08.650000+00:00"},{"alias_kind":"pith_short_8","alias_value":"3QAMPF5K","created_at":"2026-07-05T08:00:08.650000+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.15180","citing_title":"PuzzleClone: A DSL-Powered Framework for Synthesizing Verifiable Data","ref_index":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3QAMPF5KPRPJBTRBLHJNYHLGI3","json":"https://pith.science/pith/3QAMPF5KPRPJBTRBLHJNYHLGI3.json","graph_json":"https://pith.science/api/pith-number/3QAMPF5KPRPJBTRBLHJNYHLGI3/graph.json","events_json":"https://pith.science/api/pith-number/3QAMPF5KPRPJBTRBLHJNYHLGI3/events.json","paper":"https://pith.science/paper/3QAMPF5K"},"agent_actions":{"view_html":"https://pith.science/pith/3QAMPF5KPRPJBTRBLHJNYHLGI3","download_json":"https://pith.science/pith/3QAMPF5KPRPJBTRBLHJNYHLGI3.json","view_paper":"https://pith.science/paper/3QAMPF5K","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.15938&json=true","fetch_graph":"https://pith.science/api/pith-number/3QAMPF5KPRPJBTRBLHJNYHLGI3/graph.json","fetch_events":"https://pith.science/api/pith-number/3QAMPF5KPRPJBTRBLHJNYHLGI3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3QAMPF5KPRPJBTRBLHJNYHLGI3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3QAMPF5KPRPJBTRBLHJNYHLGI3/action/storage_attestation","attest_author":"https://pith.science/pith/3QAMPF5KPRPJBTRBLHJNYHLGI3/action/author_attestation","sign_citation":"https://pith.science/pith/3QAMPF5KPRPJBTRBLHJNYHLGI3/action/citation_signature","submit_replication":"https://pith.science/pith/3QAMPF5KPRPJBTRBLHJNYHLGI3/action/replication_record"}},"created_at":"2026-07-05T08:00:08.650000+00:00","updated_at":"2026-07-05T08:00:08.650000+00:00"}