{"id":"W4415622293","doi":"10.1145/3773285","title":"TaskEval: Assessing Difficulty of Code Generation Tasks for Large Language Models","year":2025,"lang":"en","type":"article","venue":"ACM Transactions on Software Engineering and Methodology","topic":"Topic Modeling","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal; Huawei Technologies (Canada)","funders":"","keywords":"Benchmarking; Benchmark (surveying); Task (project management); Code (set theory); Program comprehension; Code generation; Source code","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01312519,0.00209945,0.000896756,0.00411508,0.0005939656,0.002476002,0.001713946,0.001649793,0.003131385],"category_scores_gemma":[0.1145032,0.0004793836,0.001239851,0.002312798,0.0007499501,0.003877906,0.003689146,0.001613404,0.001622588],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001006442,"about_ca_system_score_gemma":0.001335377,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002473294,"about_ca_topic_score_gemma":0.003308526,"domain_scores_codex":[0.9846506,0.008590813,0.001825429,0.001725254,0.00280406,0.0004037205],"domain_scores_gemma":[0.8590377,0.1122489,0.00992821,0.00926369,0.007584082,0.001937486],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.004863173,0.003782401,0.1691269,0.006162766,0.00112814,0.0004591838,0.01101167,0.05393648,0.02888154,0.007765639,0.05835696,0.6545252],"study_design_scores_gemma":[0.001221494,0.009038024,0.308124,0.0008278027,0.0005115754,0.00140002,0.006549014,0.5203095,0.06526015,0.02930877,0.05660538,0.0008443152],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7226199,0.00119545,0.2317921,0.0004851197,0.0002981583,0.002275503,0.01016492,0.02154978,0.009619069],"genre_scores_gemma":[0.7841812,0.0002577919,0.1892672,0.0002227002,0.00007169814,0.003023639,0.01855018,0.001898363,0.002527113],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01312519,"threshold_uncertainty_score":0.06941348,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1129611490242461,"score_gpt":0.3607796288867569,"score_spread":0.2478184798625108,"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."}}