{"id":"W4411120155","doi":"10.18653/v1/2025.findings-naacl.410","title":"GeoCoder: Solving Geometry Problems by Generating Modular Code through Vision-Language Models","year":2025,"lang":"en","type":"article","venue":"","topic":"Model-Driven Software Engineering Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Samsung; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Modular design; Computer science; Programming language; Code (set theory); Code generation; Artificial intelligence","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":[0.0008714438,0.002305487,0.0009873139,0.001323709,0.0004815267,0.001550752,0.004906171,0.001765513,0.008559783],"category_scores_gemma":[0.005292491,0.0006298099,0.002064211,0.0008776585,0.001415693,0.003419908,0.002816669,0.00222197,0.006144204],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001337403,"about_ca_system_score_gemma":0.001957515,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008307858,"about_ca_topic_score_gemma":0.0146505,"domain_scores_codex":[0.9990265,0.0001304465,0.00006214558,0.0004023666,0.0002891106,0.00008934706],"domain_scores_gemma":[0.9985551,0.0004911711,0.0001175464,0.0005472352,0.0002259603,0.00006307111],"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":[0.0004058892,0.0002933695,0.00392193,0.001052965,0.000136396,0.0004296236,0.0004985409,0.1083574,0.03271871,0.02340533,0.09259889,0.736181],"study_design_scores_gemma":[0.0001021807,0.0001401085,0.000555141,0.00005173341,0.00002933966,0.0002352032,0.0001341751,0.9186103,0.03285114,0.01952026,0.0277195,0.00005088737],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02558633,0.0005744561,0.827375,0.0003812001,0.0001673413,0.0003869758,0.002814148,0.138149,0.004565535],"genre_scores_gemma":[0.1404376,0.0002892606,0.8383908,0.0004914796,0.00005603521,0.0004705597,0.009791093,0.005756339,0.004316855],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008559783,"threshold_uncertainty_score":0.02863538,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01027966017072345,"score_gpt":0.2580396419478833,"score_spread":0.2477599817771598,"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."}}