{"id":"W4411449683","doi":"10.1145/3729343","title":"VLATest: Testing and Evaluating Vision-Language-Action Models for Robotic Manipulation","year":2025,"lang":"en","type":"article","venue":"Proceedings of the ACM on software engineering.","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Robustness (evolution); Artificial intelligence; Software deployment; Machine learning; Generative grammar; Human–computer interaction; Software engineering","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":[],"consensus_categories":[],"category_scores_codex":[0.0003341027,0.0001247346,0.0001299166,0.0001314843,0.0001473327,0.00009102109,0.0008393736,0.00005126733,3.247624e-7],"category_scores_gemma":[0.00822916,0.0001060837,0.00004052634,0.0004803903,0.00001329015,0.0002913797,0.000457003,0.0001536534,6.639933e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005007517,"about_ca_system_score_gemma":0.00001861114,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003381345,"about_ca_topic_score_gemma":2.483131e-7,"domain_scores_codex":[0.9991795,0.000003616388,0.0002052473,0.0002874393,0.0001733008,0.0001508748],"domain_scores_gemma":[0.9986064,0.0006514203,0.0001412502,0.0003530825,0.0002192228,0.0000285647],"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":[0.000008842745,0.00003466045,0.008937951,0.0003947369,0.0000215305,3.942682e-8,0.0003803559,0.8999712,0.02028123,0.01993934,0.0001073667,0.04992276],"study_design_scores_gemma":[0.0001737395,0.00005467214,0.05086802,0.0002057838,0.0000143989,0.000002353357,0.00000928476,0.9389523,0.002071254,0.007552226,0.000006464294,0.00008953584],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3871718,0.00005767231,0.6110937,0.0007001637,0.0001314121,0.0004582741,7.020895e-7,0.0003414967,0.00004478878],"genre_scores_gemma":[0.5890462,5.454526e-7,0.4108059,0.00003364549,0.00001758397,0.0000572376,4.038546e-7,0.000009129562,0.00002933257],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.2018744,"threshold_uncertainty_score":0.9851667,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04418603467234083,"score_gpt":0.3291355321729803,"score_spread":0.2849494975006395,"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."}}