{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:KGQEOHWW7G4BLE5CL5VIVB5Z45","short_pith_number":"pith:KGQEOHWW","schema_version":"1.0","canonical_sha256":"51a0471ed6f9b81593a25f6a8a87b9e749d4eb28b183b3b4a73faded7336f76c","source":{"kind":"arxiv","id":"2408.01346","version":1},"attestation_state":"computed","paper":{"title":"Prompt Refinement or Fine-tuning? Best Practices for using LLMs in Computational Social Science Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","physics.soc-ph"],"primary_cat":"cs.CY","authors_text":"Anders Giovanni M{\\o}ller, Luca Maria Aiello","submitted_at":"2024-08-02T15:46:36Z","abstract_excerpt":"Large Language Models are expressive tools that enable complex tasks of text understanding within Computational Social Science. Their versatility, while beneficial, poses a barrier for establishing standardized best practices within the field. To bring clarity on the values of different strategies, we present an overview of the performance of modern LLM-based classification methods on a benchmark of 23 social knowledge tasks. Our results point to three best practices: select models with larger vocabulary and pre-training corpora; avoid simple zero-shot in favor of AI-enhanced prompting; fine-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":"2408.01346","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CY","submitted_at":"2024-08-02T15:46:36Z","cross_cats_sorted":["cs.CL","physics.soc-ph"],"title_canon_sha256":"5bfbb8af305e7effbf876e98f840d53a793c455bbb14db507c8ab39dd1297653","abstract_canon_sha256":"6788f69d00cdcc62939c132509ea641dafcf0748f3f1e272bcaca43c5542a447"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:51:30.476984Z","signature_b64":"mboJCLGDQizdktU0MvB5RuX9CrqKGT/h80Mlu5Kmt9CHc7e0zzyDmAjTJIypOto4eBZflJhdOuweBLrwkl/1DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"51a0471ed6f9b81593a25f6a8a87b9e749d4eb28b183b3b4a73faded7336f76c","last_reissued_at":"2026-07-05T08:51:30.476566Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:51:30.476566Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Prompt Refinement or Fine-tuning? Best Practices for using LLMs in Computational Social Science Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","physics.soc-ph"],"primary_cat":"cs.CY","authors_text":"Anders Giovanni M{\\o}ller, Luca Maria Aiello","submitted_at":"2024-08-02T15:46:36Z","abstract_excerpt":"Large Language Models are expressive tools that enable complex tasks of text understanding within Computational Social Science. Their versatility, while beneficial, poses a barrier for establishing standardized best practices within the field. To bring clarity on the values of different strategies, we present an overview of the performance of modern LLM-based classification methods on a benchmark of 23 social knowledge tasks. Our results point to three best practices: select models with larger vocabulary and pre-training corpora; avoid simple zero-shot in favor of AI-enhanced prompting; fine-t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.01346","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/2408.01346/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":"2408.01346","created_at":"2026-07-05T08:51:30.476620+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.01346v1","created_at":"2026-07-05T08:51:30.476620+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.01346","created_at":"2026-07-05T08:51:30.476620+00:00"},{"alias_kind":"pith_short_12","alias_value":"KGQEOHWW7G4B","created_at":"2026-07-05T08:51:30.476620+00:00"},{"alias_kind":"pith_short_16","alias_value":"KGQEOHWW7G4BLE5C","created_at":"2026-07-05T08:51:30.476620+00:00"},{"alias_kind":"pith_short_8","alias_value":"KGQEOHWW","created_at":"2026-07-05T08:51:30.476620+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.07368","citing_title":"Extracting Participation in Collective Action from Social Media","ref_index":43,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KGQEOHWW7G4BLE5CL5VIVB5Z45","json":"https://pith.science/pith/KGQEOHWW7G4BLE5CL5VIVB5Z45.json","graph_json":"https://pith.science/api/pith-number/KGQEOHWW7G4BLE5CL5VIVB5Z45/graph.json","events_json":"https://pith.science/api/pith-number/KGQEOHWW7G4BLE5CL5VIVB5Z45/events.json","paper":"https://pith.science/paper/KGQEOHWW"},"agent_actions":{"view_html":"https://pith.science/pith/KGQEOHWW7G4BLE5CL5VIVB5Z45","download_json":"https://pith.science/pith/KGQEOHWW7G4BLE5CL5VIVB5Z45.json","view_paper":"https://pith.science/paper/KGQEOHWW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.01346&json=true","fetch_graph":"https://pith.science/api/pith-number/KGQEOHWW7G4BLE5CL5VIVB5Z45/graph.json","fetch_events":"https://pith.science/api/pith-number/KGQEOHWW7G4BLE5CL5VIVB5Z45/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KGQEOHWW7G4BLE5CL5VIVB5Z45/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KGQEOHWW7G4BLE5CL5VIVB5Z45/action/storage_attestation","attest_author":"https://pith.science/pith/KGQEOHWW7G4BLE5CL5VIVB5Z45/action/author_attestation","sign_citation":"https://pith.science/pith/KGQEOHWW7G4BLE5CL5VIVB5Z45/action/citation_signature","submit_replication":"https://pith.science/pith/KGQEOHWW7G4BLE5CL5VIVB5Z45/action/replication_record"}},"created_at":"2026-07-05T08:51:30.476620+00:00","updated_at":"2026-07-05T08:51:30.476620+00:00"}