{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UTKIMQGGL5W2HOC4EYGWEHTINZ","short_pith_number":"pith:UTKIMQGG","schema_version":"1.0","canonical_sha256":"a4d48640c65f6da3b85c260d621e686e60c32eb00695b86915aaa7003d63ef1b","source":{"kind":"arxiv","id":"2605.26266","version":1},"attestation_state":"computed","paper":{"title":"Quantized Keys Steal Attention: Bias Correction for KV-Cache Compression in Video Diffusion","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.GR","eess.IV"],"primary_cat":"cs.LG","authors_text":"Felix Becker, Thomas Pfeil, Tuna Tuncer","submitted_at":"2026-05-25T18:51:59Z","abstract_excerpt":"Chunk-wise autoregressive video diffusion models rely on a KV cache of previously generated chunks to avoid redundant computation, but this cache quickly becomes a memory bottleneck as videos grow longer. Methods that quantize the KV cache to low bitwidths reduce memory pressure but degrade video quality. We show that a key driver of this degradation is a systematic bias in attention weights: due to the convexity of the exponential in softmax attention, quantization noise inflates the contribution of cached keys, a phenomenon we call the Jensen bias. This effect causes quantized keys to steal "},"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":"2605.26266","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-25T18:51:59Z","cross_cats_sorted":["cs.AI","cs.CV","cs.GR","eess.IV"],"title_canon_sha256":"6749c08b8eb678c2a2782971fafe98f303b4c98f2094f4857e01b7dd866dfb33","abstract_canon_sha256":"adfd8f7d13226e4a78d662e5ed46c0ea6f7b0ae21ac53745f1f7fba8d5f69321"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T01:05:09.040818Z","signature_b64":"su6pmZuBdgmB92UZkEKRho8Xdv82Nj+4gU3gSkzGIJNIik42n/n1IKdBKLuEHjYjtKNVyyrWr7bru3t0vupUDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a4d48640c65f6da3b85c260d621e686e60c32eb00695b86915aaa7003d63ef1b","last_reissued_at":"2026-05-27T01:05:09.039966Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T01:05:09.039966Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Quantized Keys Steal Attention: Bias Correction for KV-Cache Compression in Video Diffusion","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.GR","eess.IV"],"primary_cat":"cs.LG","authors_text":"Felix Becker, Thomas Pfeil, Tuna Tuncer","submitted_at":"2026-05-25T18:51:59Z","abstract_excerpt":"Chunk-wise autoregressive video diffusion models rely on a KV cache of previously generated chunks to avoid redundant computation, but this cache quickly becomes a memory bottleneck as videos grow longer. Methods that quantize the KV cache to low bitwidths reduce memory pressure but degrade video quality. We show that a key driver of this degradation is a systematic bias in attention weights: due to the convexity of the exponential in softmax attention, quantization noise inflates the contribution of cached keys, a phenomenon we call the Jensen bias. This effect causes quantized keys to steal "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.26266","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/2605.26266/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":"2605.26266","created_at":"2026-05-27T01:05:09.040100+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.26266v1","created_at":"2026-05-27T01:05:09.040100+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.26266","created_at":"2026-05-27T01:05:09.040100+00:00"},{"alias_kind":"pith_short_12","alias_value":"UTKIMQGGL5W2","created_at":"2026-05-27T01:05:09.040100+00:00"},{"alias_kind":"pith_short_16","alias_value":"UTKIMQGGL5W2HOC4","created_at":"2026-05-27T01:05:09.040100+00:00"},{"alias_kind":"pith_short_8","alias_value":"UTKIMQGG","created_at":"2026-05-27T01:05:09.040100+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/UTKIMQGGL5W2HOC4EYGWEHTINZ","json":"https://pith.science/pith/UTKIMQGGL5W2HOC4EYGWEHTINZ.json","graph_json":"https://pith.science/api/pith-number/UTKIMQGGL5W2HOC4EYGWEHTINZ/graph.json","events_json":"https://pith.science/api/pith-number/UTKIMQGGL5W2HOC4EYGWEHTINZ/events.json","paper":"https://pith.science/paper/UTKIMQGG"},"agent_actions":{"view_html":"https://pith.science/pith/UTKIMQGGL5W2HOC4EYGWEHTINZ","download_json":"https://pith.science/pith/UTKIMQGGL5W2HOC4EYGWEHTINZ.json","view_paper":"https://pith.science/paper/UTKIMQGG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.26266&json=true","fetch_graph":"https://pith.science/api/pith-number/UTKIMQGGL5W2HOC4EYGWEHTINZ/graph.json","fetch_events":"https://pith.science/api/pith-number/UTKIMQGGL5W2HOC4EYGWEHTINZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UTKIMQGGL5W2HOC4EYGWEHTINZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UTKIMQGGL5W2HOC4EYGWEHTINZ/action/storage_attestation","attest_author":"https://pith.science/pith/UTKIMQGGL5W2HOC4EYGWEHTINZ/action/author_attestation","sign_citation":"https://pith.science/pith/UTKIMQGGL5W2HOC4EYGWEHTINZ/action/citation_signature","submit_replication":"https://pith.science/pith/UTKIMQGGL5W2HOC4EYGWEHTINZ/action/replication_record"}},"created_at":"2026-05-27T01:05:09.040100+00:00","updated_at":"2026-05-27T01:05:09.040100+00:00"}