{"public_id":"co_cd308a7254a3de63a51b04cb366a61b3","status":"active","merged_into_public_id":null,"resolved_public_id":"co_cd308a7254a3de63a51b04cb366a61b3","name":"baselines","description":"Reference methods used for comparison against the proposed fine-tuning approach.","aliases":[],"types":["baseline"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"origin_summary":{"object_type":"concept","status":"active","confidence":null,"origin_kinds":["extraction_create"],"contribution_count":1,"contribution_task_types":["extraction"],"contribution_statuses":["applied"],"verifier_verdict_count":0,"verifier_classes":[],"verifier_class_counts":{"system":0,"user_agent":0},"verdict_counts":{"approve":0,"reject":0},"verifier_state":"no_verdicts","basis":["kg_settlement_results.decision_payload.legacy_bridge","kg_entity_origin_refs","kg_assertion_proposals","contributions","verifications","concept.status"],"limits":["ledger provenance is aggregated; raw contribution and verifier audit rows are not expanded","entity matching uses settlement bridge refs and edge commands"]},"papers":[{"corpus_id":276575843,"title":"Language Model Fine-Tuning on Scaled Survey Data for Predicting Distributions of Public Opinions","citation_count":35,"url":"https://sah.borca.ai/papers/276575843"}],"claims":[{"public_id":"cl_60bfcaa0556f416093dd346e9acfb695","text":"Fine-tuning on SubPOP reduces the LLM-human gap by up to 46% compared with baselines.","corpus_id":276575843,"url":"https://sah.borca.ai/claims/cl_60bfcaa0556f416093dd346e9acfb695"}],"related_concepts":[],"resolved_url":"https://sah.borca.ai/concepts/co_cd308a7254a3de63a51b04cb366a61b3","url":"https://sah.borca.ai/concepts/co_cd308a7254a3de63a51b04cb366a61b3"}