{"id":"W4411688397","doi":"10.1109/raise66696.2025.00006","title":"Towards the LLM-Based Generation of Formal Specifications from Natural-Language Contracts: Early Experiments with Symboleo","year":2025,"lang":"en","type":"article","venue":"","topic":"Artificial Intelligence in Law","field":"Social Sciences","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"","keywords":"Computer science; Programming language; Natural language generation; Natural language; Formal methods; Natural (archaeology); Software engineering; Natural language processing; History","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":[0.0002781198,0.0000793818,0.0001045377,0.00004118882,0.0004509734,0.00009964828,0.0003080418,0.00005748382,0.0004552655],"category_scores_gemma":[0.0001162859,0.00005114863,0.00004247203,0.0002816945,0.0003842464,0.0003011069,0.00001854698,0.00009267047,0.00002109498],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000691898,"about_ca_system_score_gemma":0.0002453679,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01544536,"about_ca_topic_score_gemma":0.009564761,"domain_scores_codex":[0.9990223,0.0001073831,0.0002068888,0.0001537029,0.0003179187,0.0001918183],"domain_scores_gemma":[0.9992934,0.0001658589,0.00009022544,0.0002431429,0.0001727635,0.00003463265],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0001414711,0.0003858015,0.003080857,0.000004562199,0.0001254686,0.000003045492,0.0625556,0.0002642368,0.1003804,0.785884,0.002461751,0.04471282],"study_design_scores_gemma":[0.000294642,0.00009782545,0.01130529,0.00004961072,0.00005664804,7.345579e-8,0.05772283,0.004956883,0.9113396,0.001341956,0.0125816,0.0002530088],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8884825,0.0003066091,0.03086219,0.004124933,0.0004308432,0.0004927229,0.00001433508,0.00006560265,0.07522032],"genre_scores_gemma":[0.9961983,0.000007394778,0.001627584,0.0004133152,0.0001598025,0.00003539301,0.00001446662,0.000004646884,0.001539072],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8109592,"threshold_uncertainty_score":0.9911109,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09567948888282997,"score_gpt":0.3632091950845656,"score_spread":0.2675297062017357,"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."}}