{"id":"W6979343253","doi":"","title":"Can LLMs Reason Abstractly Over Math Word Problems Without CoT? Disentangling Abstract Formulation From Arithmetic Computation","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":"Mitacs; Canadian Institute for Advanced Research; Nvidia","keywords":"Conflation; Computation; Causal reasoning; Word (group theory); Automated reasoning; Mental arithmetic; Multiplication (music)","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":[0.005298142,0.001091002,0.0009119962,0.001059075,0.0005753233,0.005095216,0.002054569,0.001461822,0.01002743],"category_scores_gemma":[0.04829535,0.0006442177,0.001832775,0.0008879835,0.002212075,0.01356307,0.004215698,0.003333864,0.00294879],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001854536,"about_ca_system_score_gemma":0.003126847,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005526968,"about_ca_topic_score_gemma":0.0104344,"domain_scores_codex":[0.9956406,0.001827924,0.0003030929,0.0007038289,0.001229782,0.0002947204],"domain_scores_gemma":[0.9772613,0.0134204,0.001010127,0.005601205,0.002115482,0.0005914601],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.00147751,0.0005855209,0.03469262,0.001866211,0.0003608713,0.0004544382,0.004531734,0.1939492,0.02657767,0.3443074,0.02303188,0.368165],"study_design_scores_gemma":[0.00006635393,0.0002218824,0.002373902,0.0001528762,0.0001107043,0.0001333403,0.0008789994,0.6034605,0.01817114,0.3559725,0.01838416,0.00007375427],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2623989,0.0006995482,0.6920386,0.003491123,0.0002727769,0.0002305519,0.002190792,0.01631487,0.02236297],"genre_scores_gemma":[0.7231978,0.000246443,0.2667153,0.0006232532,0.00005030657,0.0001710901,0.002832836,0.001913451,0.004249466],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01002743,"threshold_uncertainty_score":0.03354514,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03054045886540415,"score_gpt":0.275435382404612,"score_spread":0.2448949235392078,"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."}}