{"id":"W6979293534","doi":"","title":"Syntactic and Semantic Control of Large Language Models via Sequential Monte Carlo","year":2025,"lang":"en","type":"article","venue":"ArXiv.org","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Canadian Institute for Advanced Research","keywords":"Probabilistic logic; Python (programming language); Inference; Language model; Variety (cybernetics); Range (aeronautics); Monte Carlo method","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":[0.004338199,0.000899365,0.0007562191,0.0006708954,0.0007318137,0.001869336,0.002527023,0.001124604,0.005704565],"category_scores_gemma":[0.01776571,0.0008121549,0.001361253,0.0005349447,0.001815878,0.002434489,0.002364329,0.002464789,0.001628379],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001567332,"about_ca_system_score_gemma":0.00208409,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00536479,"about_ca_topic_score_gemma":0.01020198,"domain_scores_codex":[0.9975237,0.001282614,0.0001157055,0.0004410398,0.0005109739,0.0001259231],"domain_scores_gemma":[0.9907726,0.007284303,0.0002968438,0.0009639179,0.0005260811,0.0001563641],"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":[0.0002929795,0.0001567219,0.002187994,0.0002828912,0.0001294305,0.0002378175,0.0005301112,0.7559004,0.00998241,0.1238674,0.005999656,0.1004321],"study_design_scores_gemma":[0.00002485602,0.00001040146,0.00004184751,0.000006582002,0.000006960923,0.0000130845,0.00001082181,0.9647844,0.001860719,0.03167279,0.001561543,0.000005942511],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009463482,0.00008362335,0.9823217,0.000249747,0.00003099697,0.00008997438,0.0001265799,0.005792139,0.001841698],"genre_scores_gemma":[0.3860916,0.0001549636,0.606569,0.0003667713,0.00007902139,0.0005955188,0.0006970384,0.00245416,0.002991955],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005704565,"threshold_uncertainty_score":0.0229429,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0179119072032372,"score_gpt":0.2600284982508983,"score_spread":0.2421165910476611,"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."}}