{"id":"W4402683916","doi":"10.18653/v1/2024.acl-srw.35","title":"Trace-of-Thought Prompting: Investigating Prompt-Based Knowledge Distillation Through Question Decomposition","year":2024,"lang":"en","type":"article","venue":"","topic":"Intelligent Tutoring Systems and Adaptive Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"TRACE (psycholinguistics); Decomposition; Computer science; Distillation; Artificial intelligence; Chemistry; Linguistics; Chromatography; Philosophy; Organic chemistry","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":[0.004642655,0.0007170922,0.0006211557,0.0007227162,0.0003904401,0.002460534,0.001630018,0.001300625,0.007371147],"category_scores_gemma":[0.1078554,0.0005364586,0.0003328136,0.0006074836,0.0009065886,0.005559049,0.002915377,0.002003543,0.0008803873],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00055993,"about_ca_system_score_gemma":0.001200283,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001700039,"about_ca_topic_score_gemma":0.001208153,"domain_scores_codex":[0.9966574,0.001439516,0.0001810381,0.0007968453,0.0007636285,0.0001616123],"domain_scores_gemma":[0.9277963,0.05828106,0.004430669,0.005189415,0.002980703,0.001321804],"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":[0.01111808,0.002240523,0.05171216,0.003299252,0.0002610236,0.0009973815,0.04572816,0.01528923,0.3197961,0.03643389,0.003213932,0.5099102],"study_design_scores_gemma":[0.001379373,0.006150756,0.1451154,0.0008976144,0.0007013988,0.002110593,0.01240562,0.481884,0.1763652,0.1451487,0.02746688,0.0003745852],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6581355,0.0005491069,0.3224671,0.000538855,0.0001915266,0.001205517,0.0009419352,0.00284455,0.01312596],"genre_scores_gemma":[0.9313659,0.0001519893,0.06499334,0.0002054888,0.0000303965,0.000366092,0.0006501246,0.0003197193,0.001916912],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.007371147,"threshold_uncertainty_score":0.02465898,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03695600329241108,"score_gpt":0.3365478367036141,"score_spread":0.299591833411203,"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."}}