{"id":"W4401183144","doi":"10.26434/chemrxiv-2024-mm31v","title":"Assessment of Fine-Tuned Large Language Models for Real-World Chemistry and Material Science Applications","year":2024,"lang":"en","type":"preprint","venue":"ChemRxiv","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; University of Waterloo","funders":"","keywords":"Benchmark (surveying); Set (abstract data type); Computer science; Range (aeronautics); Computation; Work (physics); Simple (philosophy); Fine-tuning; Machine learning; Artificial intelligence; Algorithm; Materials science; Epistemology; Engineering; Geography","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007388203,0.001789915,0.001165749,0.00111613,0.0006957726,0.001449554,0.002723726,0.003007631,0.002323375],"category_scores_gemma":[0.02554983,0.0007470798,0.001262369,0.000879564,0.001035448,0.002664505,0.001436349,0.003364354,0.001159444],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002014675,"about_ca_system_score_gemma":0.001416458,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01111686,"about_ca_topic_score_gemma":0.01221788,"domain_scores_codex":[0.9981584,0.0009732963,0.00008719784,0.0004271061,0.0002215106,0.0001324273],"domain_scores_gemma":[0.9847467,0.0120835,0.0004093956,0.001498028,0.0008856711,0.000376769],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0006692781,0.0005511593,0.004348989,0.0003126583,0.0002432771,0.0001052397,0.00008819092,0.950789,0.001985389,0.002129997,0.006748423,0.03202836],"study_design_scores_gemma":[0.00006390512,0.0001100403,0.0004850031,0.00001807967,0.00002048259,0.00001967158,0.00002764039,0.994752,0.001464951,0.002317567,0.0007063649,0.00001414083],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8372283,0.005627132,0.1185376,0.004352327,0.0006302645,0.0004075432,0.007610149,0.01481413,0.01079244],"genre_scores_gemma":[0.9375111,0.000453986,0.04960788,0.000729189,0.0001020954,0.0003083984,0.008930747,0.0005403896,0.001816142],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01111686,"threshold_uncertainty_score":0.03907305,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01532537674670113,"score_gpt":0.3398288895007431,"score_spread":0.324503512754042,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}