{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4RAP65AD6F4W4FX67ZXJZVQRIW","short_pith_number":"pith:4RAP65AD","schema_version":"1.0","canonical_sha256":"e440ff7403f1796e16fefe6e9cd61145973c58eb0ca4cd8e5782998e9f65327c","source":{"kind":"arxiv","id":"2607.13205","version":1},"attestation_state":"computed","paper":{"title":"Adaptive Filtering of the KV Cache: Diagnosing and Correcting Structural-Role Bias in LLM Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Soumil Mandal","submitted_at":"2026-07-14T18:55:20Z","abstract_excerpt":"Attention-based KV cache eviction (H2O and its descendants) compresses the memory-constrained state of a long-context model by ranking tokens on accumulated attention mass, treated here as signal energy, and keeping the heaviest. On schema-dense input streams such as nested JSON, this score acts as a non-stationary filter that disproportionately retains noise: a non-content sink role (delimiters or whitespace) carries an order of magnitude more energy than any content role, and structural KEY tokens are over-retained at roughly 1.8x the rate of the answer-carrying VALUE tokens, collapsing exac"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2607.13205","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-07-14T18:55:20Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"21a4c8fdaee4ef3b4b6fef27950013dcfd3e70f13d2ff485b44d7eff1939ce83","abstract_canon_sha256":"6ffef4cf6b8cda535c2e1acea2215139119f4bfd482bb4139501d2e2c4f5d137"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-16T00:22:02.778681Z","signature_b64":"LkizmNRzm6VdfNhT1wbye8WsUQ5WnmyweyvMUKq+5IaylGOVBGFJv7p4abpydbMfcYkxUMRSDmu+BzwpPT6/Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e440ff7403f1796e16fefe6e9cd61145973c58eb0ca4cd8e5782998e9f65327c","last_reissued_at":"2026-07-16T00:22:02.777829Z","signature_status":"signed_v1","first_computed_at":"2026-07-16T00:22:02.777829Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Adaptive Filtering of the KV Cache: Diagnosing and Correcting Structural-Role Bias in LLM Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Soumil Mandal","submitted_at":"2026-07-14T18:55:20Z","abstract_excerpt":"Attention-based KV cache eviction (H2O and its descendants) compresses the memory-constrained state of a long-context model by ranking tokens on accumulated attention mass, treated here as signal energy, and keeping the heaviest. On schema-dense input streams such as nested JSON, this score acts as a non-stationary filter that disproportionately retains noise: a non-content sink role (delimiters or whitespace) carries an order of magnitude more energy than any content role, and structural KEY tokens are over-retained at roughly 1.8x the rate of the answer-carrying VALUE tokens, collapsing exac"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.13205","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2607.13205/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2607.13205","created_at":"2026-07-16T00:22:02.778270+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.13205v1","created_at":"2026-07-16T00:22:02.778270+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.13205","created_at":"2026-07-16T00:22:02.778270+00:00"},{"alias_kind":"pith_short_12","alias_value":"4RAP65AD6F4W","created_at":"2026-07-16T00:22:02.778270+00:00"},{"alias_kind":"pith_short_16","alias_value":"4RAP65AD6F4W4FX6","created_at":"2026-07-16T00:22:02.778270+00:00"},{"alias_kind":"pith_short_8","alias_value":"4RAP65AD","created_at":"2026-07-16T00:22:02.778270+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4RAP65AD6F4W4FX67ZXJZVQRIW","json":"https://pith.science/pith/4RAP65AD6F4W4FX67ZXJZVQRIW.json","graph_json":"https://pith.science/api/pith-number/4RAP65AD6F4W4FX67ZXJZVQRIW/graph.json","events_json":"https://pith.science/api/pith-number/4RAP65AD6F4W4FX67ZXJZVQRIW/events.json","paper":"https://pith.science/paper/4RAP65AD"},"agent_actions":{"view_html":"https://pith.science/pith/4RAP65AD6F4W4FX67ZXJZVQRIW","download_json":"https://pith.science/pith/4RAP65AD6F4W4FX67ZXJZVQRIW.json","view_paper":"https://pith.science/paper/4RAP65AD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.13205&json=true","fetch_graph":"https://pith.science/api/pith-number/4RAP65AD6F4W4FX67ZXJZVQRIW/graph.json","fetch_events":"https://pith.science/api/pith-number/4RAP65AD6F4W4FX67ZXJZVQRIW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4RAP65AD6F4W4FX67ZXJZVQRIW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4RAP65AD6F4W4FX67ZXJZVQRIW/action/storage_attestation","attest_author":"https://pith.science/pith/4RAP65AD6F4W4FX67ZXJZVQRIW/action/author_attestation","sign_citation":"https://pith.science/pith/4RAP65AD6F4W4FX67ZXJZVQRIW/action/citation_signature","submit_replication":"https://pith.science/pith/4RAP65AD6F4W4FX67ZXJZVQRIW/action/replication_record"}},"created_at":"2026-07-16T00:22:02.778270+00:00","updated_at":"2026-07-16T00:22:02.778270+00:00"}