{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:RJFZVYLJ3WT5LJ3YS5EPVOKUBA","short_pith_number":"pith:RJFZVYLJ","schema_version":"1.0","canonical_sha256":"8a4b9ae169dda7d5a7789748fab954080fb01dea5a4e5ab8492a5520ab2a1381","source":{"kind":"arxiv","id":"2403.12776","version":1},"attestation_state":"computed","paper":{"title":"Automated Data Curation for Robust Language Model Fine-Tuning","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jiuhai Chen, Jonas Mueller","submitted_at":"2024-03-19T14:44:45Z","abstract_excerpt":"Large Language Models have become the de facto approach to sequence-to-sequence text generation tasks, but for specialized tasks/domains, a pretrained LLM lacks specific capabilities to produce accurate or well-formatted responses. Supervised fine-tuning specializes a LLM by training it on dataset of example prompts with target responses, but real-world data tends to be noisy. While many fine-tuning algorithms exist, here we consider a \\emph{data-centric AI} perspective on LLM fine-tuning, studying how to \\emph{systematically} curate the training dataset to improve the LLM produced via \\emph{a"},"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":"2403.12776","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-19T14:44:45Z","cross_cats_sorted":[],"title_canon_sha256":"a9fa2cc6ac743191627f511d3a76f37d6ee969092cc6343adbc9d58b11bb9f37","abstract_canon_sha256":"02430b20e06d0117ba3accafa3017e2aea5807fa8192ba460d89e190a4acc6b2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:58:06.926437Z","signature_b64":"fwyobkVCRVt6NrBtYIr9/Vq4Gc0W6gRe9QMAigG5jaV+dIi+W1Zrbhfjdem3vuauX0gVJS515KEsh373PNdADw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8a4b9ae169dda7d5a7789748fab954080fb01dea5a4e5ab8492a5520ab2a1381","last_reissued_at":"2026-07-05T07:58:06.925836Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:58:06.925836Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Automated Data Curation for Robust Language Model Fine-Tuning","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jiuhai Chen, Jonas Mueller","submitted_at":"2024-03-19T14:44:45Z","abstract_excerpt":"Large Language Models have become the de facto approach to sequence-to-sequence text generation tasks, but for specialized tasks/domains, a pretrained LLM lacks specific capabilities to produce accurate or well-formatted responses. Supervised fine-tuning specializes a LLM by training it on dataset of example prompts with target responses, but real-world data tends to be noisy. While many fine-tuning algorithms exist, here we consider a \\emph{data-centric AI} perspective on LLM fine-tuning, studying how to \\emph{systematically} curate the training dataset to improve the LLM produced via \\emph{a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.12776","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/2403.12776/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":"2403.12776","created_at":"2026-07-05T07:58:06.925904+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.12776v1","created_at":"2026-07-05T07:58:06.925904+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.12776","created_at":"2026-07-05T07:58:06.925904+00:00"},{"alias_kind":"pith_short_12","alias_value":"RJFZVYLJ3WT5","created_at":"2026-07-05T07:58:06.925904+00:00"},{"alias_kind":"pith_short_16","alias_value":"RJFZVYLJ3WT5LJ3Y","created_at":"2026-07-05T07:58:06.925904+00:00"},{"alias_kind":"pith_short_8","alias_value":"RJFZVYLJ","created_at":"2026-07-05T07:58:06.925904+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.09665","citing_title":"Learning Multi-Indicator Weights for Data Selection: A Joint Task-Model Adaptation Framework with Efficient Proxies","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19444","citing_title":"Unsupervised Confidence Calibration for Reasoning LLMs from a Single Generation","ref_index":88,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RJFZVYLJ3WT5LJ3YS5EPVOKUBA","json":"https://pith.science/pith/RJFZVYLJ3WT5LJ3YS5EPVOKUBA.json","graph_json":"https://pith.science/api/pith-number/RJFZVYLJ3WT5LJ3YS5EPVOKUBA/graph.json","events_json":"https://pith.science/api/pith-number/RJFZVYLJ3WT5LJ3YS5EPVOKUBA/events.json","paper":"https://pith.science/paper/RJFZVYLJ"},"agent_actions":{"view_html":"https://pith.science/pith/RJFZVYLJ3WT5LJ3YS5EPVOKUBA","download_json":"https://pith.science/pith/RJFZVYLJ3WT5LJ3YS5EPVOKUBA.json","view_paper":"https://pith.science/paper/RJFZVYLJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.12776&json=true","fetch_graph":"https://pith.science/api/pith-number/RJFZVYLJ3WT5LJ3YS5EPVOKUBA/graph.json","fetch_events":"https://pith.science/api/pith-number/RJFZVYLJ3WT5LJ3YS5EPVOKUBA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RJFZVYLJ3WT5LJ3YS5EPVOKUBA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RJFZVYLJ3WT5LJ3YS5EPVOKUBA/action/storage_attestation","attest_author":"https://pith.science/pith/RJFZVYLJ3WT5LJ3YS5EPVOKUBA/action/author_attestation","sign_citation":"https://pith.science/pith/RJFZVYLJ3WT5LJ3YS5EPVOKUBA/action/citation_signature","submit_replication":"https://pith.science/pith/RJFZVYLJ3WT5LJ3YS5EPVOKUBA/action/replication_record"}},"created_at":"2026-07-05T07:58:06.925904+00:00","updated_at":"2026-07-05T07:58:06.925904+00:00"}