{"id":"W4411120204","doi":"10.18653/v1/2025.findings-naacl.424","title":"Tooling or Not Tooling? The Impact of Tools on Language Agents for Chemistry Problem Solving","year":2025,"lang":"en","type":"article","venue":"","topic":"Semantic Web and Ontologies","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Argonne National Laboratory; U.S. National Library of Medicine; Institute for Catastrophic Loss Reduction; National Science Foundation","keywords":"Computer science; Manufacturing engineering; Software engineering; Process engineering; Engineering","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01105355,0.00143787,0.0008746189,0.0008058317,0.0005702187,0.003171125,0.0023924,0.001736981,0.003769431],"category_scores_gemma":[0.042746,0.0008457743,0.001246067,0.0006308158,0.001861871,0.01004468,0.004818731,0.003769924,0.001380626],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00103925,"about_ca_system_score_gemma":0.002259632,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00432406,"about_ca_topic_score_gemma":0.005699542,"domain_scores_codex":[0.9907674,0.00554302,0.0006820203,0.001075108,0.001517301,0.0004150036],"domain_scores_gemma":[0.9514335,0.03601639,0.001536738,0.008206105,0.001833393,0.0009739378],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.003254835,0.001960174,0.01906552,0.003083014,0.0006987738,0.000515828,0.002103817,0.2574427,0.03183576,0.07411554,0.01978396,0.5861401],"study_design_scores_gemma":[0.0005000009,0.0009101534,0.002192438,0.0003843306,0.0003810579,0.0002262479,0.0004735485,0.8535336,0.03246599,0.04238769,0.06641247,0.0001325908],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3832217,0.005760511,0.5184478,0.006972246,0.0006202661,0.0005613575,0.0009007278,0.05047385,0.03304157],"genre_scores_gemma":[0.5314404,0.001101828,0.4601873,0.001041294,0.00006498111,0.0002079368,0.001033213,0.001758526,0.003164418],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01105355,"threshold_uncertainty_score":0.05845743,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04620901255201522,"score_gpt":0.3422979205710734,"score_spread":0.2960889080190582,"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."}}