{"paper":{"title":"Page-EntroKV: Hardware-Aligned, Entropy-Weighted KV-Cache Eviction under Grouped-Query Attention","arxiv_id":"2610.03135","primary_cat":"cs.LG","submitted_at":"2026-10-02T11:01:39+00:00","version_compiled":"v1"},"notice":"Every quoted sentence is the authors' text from the arXiv listing abstract. Labels and probabilities are the outputs of classifier-1.13.0 on the question bank abstract-p1-20260923, exactly as returned. No sentence in this record was written by a language model.","schema":"pith.compiled-record/v0","status":"compiled","abstract":{"source":"arXiv listing abstract v1","sentences":[{"i":0,"p":1.0,"id":"s0","own":false,"role":"background","text":"Serving long-context autoregressive language models is constrained by the key-value (KV) cache.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.04,"main_claim_p":0.02,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.0,"background":1.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":0.97,"id":"s1","own":false,"role":"background","text":"Most dynamic eviction methods score token importance per query head and choose tokens independently.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.03,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.08,"role_probabilities":{"gap":0.03,"method":0.0,"result":0.0,"background":0.97,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":2,"p":0.63,"id":"s2","own":false,"role":"background","text":"This fits poorly with grouped-query attention (GQA), where several query heads share one physical KV buffer: divergent per-head selections force the serving engine to retain the union of their choices - inflating the cache by up to the group ratio r - while arithmetic mean pooling dilutes the specialized retrieval heads that carry factual recall.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.11,"main_claim_p":0.07,"quantitative":true,"quantitative_p":0.55,"role_probabilities":{"gap":0.37,"method":0.0,"result":0.0,"background":0.63,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":3,"p":0.99,"id":"s3","own":true,"role":"contribution","text":"We introduce Page-EntroKV, a formal framework for KV-cache eviction operating at the granularity GQA serving actually allocates.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.92,"main_claim_p":0.11,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.0,"method":0.01,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.99,"interpretation":0.0}},{"i":4,"p":0.99,"id":"s4","own":true,"role":"method","text":"Heads within each physical group are pooled by weights derived from sink-isolated collision (Renyi-2) entropy - one inner product per head, computed once at prefill with no calibration - so sink heads cannot masquerade as retrieval heads.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.92,"main_claim_p":0.09,"quantitative":true,"quantitative_p":0.52,"role_probabilities":{"gap":0.0,"method":0.99,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.01,"interpretation":0.0}},{"i":5,"p":0.98,"id":"s5","own":true,"role":"method","text":"Pooled scores are projected onto PagedAttention page frames, and eviction executes at the hardware tuple (layer, group, page).","hedge":{"p":0.99,"label":"asserted"},"own_p":0.85,"main_claim_p":0.06,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.0,"method":0.98,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.02,"interpretation":0.0}},{"i":6,"p":0.98,"id":"s6","own":true,"role":"contribution","text":"We formalize the union overhead ratio (UOR) and intra-group disagreement, prove an exact identity linking them for two-head groups alongside two-sided bounds at every group ratio, prove strict budget preservation and a finite-context needle-retention bound that arithmetic mean pooling provably violates, and give exact per-layer page accounting.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.96,"main_claim_p":0.2,"quantitative":true,"quantitative_p":0.85,"role_probabilities":{"gap":0.0,"method":0.01,"result":0.01,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.98,"interpretation":0.0}},{"i":7,"p":1.0,"id":"s7","own":true,"role":"result","text":"On a pilot architecture (Qwen2.5-1.5B-Instruct, r=6), head-independent replay over 2,240 group measurements yields union overhead up to 4.75x at a 2% budget, while Page-EntroKV holds UOR exactly 1.000; sink isolation removes a 13x sink masquerade; needle recall is 100% versus 0% for mean pooling at a 20% budget; retained cardinality is exact for every page size; and QA and code tasks remain solvable at 20% retention.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.95,"main_claim_p":0.59,"quantitative":true,"quantitative_p":0.99,"role_probabilities":{"gap":0.0,"method":0.0,"result":1.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}}]},"compiled":{"main_claim":{"p":0.92,"text":"On a pilot architecture (Qwen2.5-1.5B-Instruct, r=6), head-independent replay over 2,240 group measurements yields union overhead up to 4.75x at a 2% budget, while Page-EntroKV holds UOR exactly 1.000; sink isolation removes a 13x sink masquerade; needle recall is 100% versus 0% for mean pooling at a 20% budget; retained cardinality is exact for every page size; and QA and code tasks remain solvable at 20% retention.","location":"abstract s7","hedge_strength":{"p":1.0,"label":"asserted"}},"provenance":{"route":{"via":"TypeSafe API","model":"classifier-1.13.0","endpoint":"https://api.typesafe.ai/v1/systemone","provider":"TypeSafe"},"packing":"one request per paper","requests":1,"input_tokens":6660,"cost_usd_reported":0.0},"compiled_at":"2026-10-05","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.11,"states_limitation":0.4,"human_participants":0.02,"controlled_comparison":0.92,"randomized_assignment":0.02,"states_quantitative_result":0.99,"compares_with_prior_results":0.97,"mentions_code_or_data_availability":0.03},"evidential_basis":{"p":0.62,"value":"observation or experiment","probabilities":{"argument":0.0,"literature":0.0,"proof_or_derivation":0.24,"analysis_of_existing_data":0.0,"observation_or_experiment":0.62,"simulation_or_computation":0.14}},"contribution_type":{"p":0.99,"value":"method or tool","probabilities":{"review":0.0,"method_or_tool":0.99,"empirical_study":0.0,"simulation_study":0.0,"theoretical_result":0.01,"dataset_or_benchmark":0.0,"position_or_commentary":0.0}},"contribution_statement":{"p":1.0,"text":"We introduce Page-EntroKV, a formal framework for KV-cache eviction operating at the granularity GQA serving actually allocates.","location":"abstract s3"}},"coverage":{"tier":"abstract","levels_run":["P abstract"],"not_examined":["full text","figures and tables","references and citing sentences","relations to other papers"],"sources_read":["arXiv listing abstract v1"],"levels_not_run":["S sections","T body sentences","C citations","R cross-paper relations"],"sentences_examined":8,"sentences_in_abstract":8},"measured_accuracy":{"url":"https://pith.science/accuracy","measured_on":"2026-10-02","method":"two independent model families labelled a fresh random sample blind; the figure is how often the published label matches their consensus","sentence_role_agreement":0.92,"main_claim_agreement":0.908,"typed_citation_agreement":0.933,"typed_citation_agreement_95_interval":[0.917,0.945],"review_comments_accurate":0.758,"review_comments_accurate_95_interval":[0.684,0.82],"review_caught_admitted_error_found_or_related":0.448,"review_comments_judged":{"judged_by":["gpt-5.6-luna","grok-4.7"],"judged_against":"Pith's own text copy of each paper: text extracted from the PDF, the first 90,000 characters","measured_on":"2026-10-04","accurate":113,"decided":149,"sampled_reviews":200,"split_by_lane":false,"by_lane":"https://pith.science/accuracy#by-lane"}}}