{"id":"W7114912747","doi":"10.1145/3748636.3764603","title":"Learning Heuristics to Solve Dynamic Vehicle Routing Problems Using Large Language Models","year":2025,"lang":"en","type":"article","venue":"","topic":"Vehicle Routing Optimization Methods","field":"Engineering","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"National Research Council Canada; University of Toronto; University of Waterloo","funders":"National Research Council Canada","keywords":"Heuristics; Vehicle routing problem; Leverage (statistics); Scalability; Greedy algorithm; Executable; Reinforcement learning; Routing (electronic design automation)","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0006121057,0.0001783234,0.0002078776,0.0001831576,0.0001729575,0.00009073652,0.0001591178,0.00009911049,0.00003281538],"category_scores_gemma":[0.0001781221,0.0001991735,0.0000498756,0.0005996743,0.000009162361,0.0001374701,0.0001240007,0.0003005335,0.00001818215],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001781542,"about_ca_system_score_gemma":0.00002717447,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003643341,"about_ca_topic_score_gemma":0.00002406116,"domain_scores_codex":[0.9987997,0.00005484183,0.0003166846,0.0002353222,0.0001394106,0.0004539767],"domain_scores_gemma":[0.9994947,0.0001025556,0.00003129473,0.0002250929,0.00006614776,0.00008026599],"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":[0.000001550426,0.00001050956,0.0004008847,0.00006572429,0.00002345234,0.00000248851,0.001537886,0.9823591,0.01130323,0.001574779,0.00003086847,0.002689564],"study_design_scores_gemma":[0.0002413271,0.00001006311,0.0000816404,0.0001036666,0.00001944604,0.000001523811,0.000778624,0.9970084,0.001247015,0.0001868576,0.0001132277,0.0002082319],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.149039,0.0001007541,0.8408368,0.00003642633,0.000168464,0.0001907043,0.00000365031,0.0009123106,0.008711906],"genre_scores_gemma":[0.7104446,0.000005686339,0.2884875,0.00009744682,0.00002025431,0.000006311682,0.000006290495,0.00004591165,0.0008860076],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5614056,"threshold_uncertainty_score":0.8122059,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01381204195917172,"score_gpt":0.2859743451169398,"score_spread":0.2721623031577681,"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."}}