{"id":"W4409348132","doi":"10.1609/aaai.v39i23.34639","title":"Prompt Compression with Context-Aware Sentence Encoding for Fast and Improved LLM Inference","year":2025,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Encoding (memory); Sentence; Inference; Computer science; Context (archaeology); Compression (physics); Natural language processing; Artificial intelligence; Speech recognition; History","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":[0.001132032,0.001232145,0.0007127543,0.001027822,0.0003854859,0.0007768731,0.001143048,0.001040201,0.007059372],"category_scores_gemma":[0.008342328,0.0003074689,0.0007010418,0.0007237176,0.0004313062,0.002476387,0.001391629,0.002493916,0.003354752],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00070631,"about_ca_system_score_gemma":0.001235266,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002423014,"about_ca_topic_score_gemma":0.004939731,"domain_scores_codex":[0.9992651,0.0002684727,0.00005609379,0.0002076289,0.0001484222,0.0000542295],"domain_scores_gemma":[0.9978569,0.00118927,0.0001032651,0.0003940733,0.0003854604,0.00007105766],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005536999,0.0002469373,0.002090875,0.0005092771,0.00006098711,0.0002976722,0.000405127,0.0281548,0.04157463,0.008111493,0.0336311,0.8843634],"study_design_scores_gemma":[0.0001719509,0.0004047826,0.00191372,0.0001120161,0.00009637655,0.0004147027,0.0002923441,0.8776525,0.06026031,0.03500188,0.02361047,0.00006884844],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05253595,0.002647446,0.9083124,0.001319492,0.0005819116,0.0003479192,0.004379282,0.02638283,0.003492702],"genre_scores_gemma":[0.344532,0.001038539,0.631447,0.001111705,0.0004763296,0.0006024317,0.0137282,0.001055344,0.006008418],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007059372,"threshold_uncertainty_score":0.02361596,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03933237982885971,"score_gpt":0.3119017386325069,"score_spread":0.2725693588036472,"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."}}