{"id":"W4415958513","doi":"10.48550/arxiv.2510.19886","title":"An Expert-grounded benchmark of General Purpose LLMs in LCA","year":2025,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Environmental Impact and Sustainability","field":"Environmental Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal; St Mary's Hospital Centre","funders":"","keywords":"Benchmark (surveying); Quality (philosophy); Field (mathematics); Hallucinating; Usability; Sophistication; Risk assessment","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0004353341,0.0003752737,0.0004828013,0.00008710044,0.00007201031,0.00002710299,0.0007688131,0.0003648369,0.002204416],"category_scores_gemma":[0.0000660953,0.0003678906,0.0001745683,0.0002140306,0.0004377864,0.0002838692,0.001481759,0.0005324269,0.00005408713],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001036314,"about_ca_system_score_gemma":0.00005567998,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009506351,"about_ca_topic_score_gemma":0.0009630798,"domain_scores_codex":[0.9975577,0.0002202382,0.0005595874,0.0008032325,0.0003676588,0.0004915543],"domain_scores_gemma":[0.9984081,0.00004503352,0.0001740908,0.001183712,0.000006713823,0.000182305],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0000438115,0.0005477482,0.9828054,0.00005982987,0.00001415562,0.00001795778,0.001388373,0.002153879,0.009829189,0.00003016682,0.0002972149,0.00281229],"study_design_scores_gemma":[0.0003399184,0.00009829153,0.9876481,0.00004728732,0.00001420367,0.000001135966,0.0004369234,0.0005740738,0.00798646,0.001490496,0.001002063,0.0003610019],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9852799,0.0001603607,0.00006618692,0.0001620233,0.0003065138,0.0006376781,0.00003090688,0.00003480213,0.0133217],"genre_scores_gemma":[0.9962038,0.0001432094,0.000957431,0.0003338695,0.00005835289,0.00008895817,0.00008449687,0.00001933334,0.002110568],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01121113,"threshold_uncertainty_score":0.9998773,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02119877764453126,"score_gpt":0.2958000058968369,"score_spread":0.2746012282523056,"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."}}