{"id":"W4416447818","doi":"10.48550/arxiv.2505.16217","title":"Reward-Aware Proto-Representations in Reinforcement Learning","year":2025,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Alliance de recherche numérique du Canada; Alberta Innovates; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Representation (politics); Reinforcement learning; Space (punctuation); Function (biology); Key (lock); Successor cardinal; Encoding (memory)","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0006386358,0.0004643121,0.0005049309,0.0005911698,0.0002415875,0.0003061575,0.002324416,0.0003567386,0.00007592674],"category_scores_gemma":[0.0004687548,0.0005089734,0.00017489,0.0007667098,0.00007260733,0.0004057327,0.004629125,0.002073966,0.0002697135],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003853601,"about_ca_system_score_gemma":0.0005957084,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003760667,"about_ca_topic_score_gemma":0.00002333851,"domain_scores_codex":[0.9963307,0.0002676848,0.0009605174,0.001110442,0.0006676989,0.0006629514],"domain_scores_gemma":[0.9970902,0.0002147083,0.0005179347,0.001821741,0.0002279524,0.0001274824],"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":[0.000004780491,0.00002019543,0.2251031,0.0001765151,0.00004272816,0.000022971,0.0009547921,0.7716305,0.00001941083,0.00121908,0.0003497609,0.0004561915],"study_design_scores_gemma":[0.0008307583,0.0001500368,0.08476085,0.001213683,0.00003791317,0.000004038951,0.0001878689,0.9005411,0.0006249871,0.0004278279,0.01030776,0.0009131517],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02047838,0.00007041843,0.9550361,0.001283534,0.001281383,0.002273522,0.000001644694,0.0005982428,0.01897676],"genre_scores_gemma":[0.9588622,0.0001281324,0.007982063,0.0004292006,0.0001495444,0.0009864456,0.0001157275,0.0000336128,0.03131311],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.947054,"threshold_uncertainty_score":0.9997362,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05062451241968759,"score_gpt":0.318504177632027,"score_spread":0.2678796652123395,"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."}}