{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:RTVUFAF5VU5KA7MUCXRKCBJUVJ","short_pith_number":"pith:RTVUFAF5","canonical_record":{"source":{"id":"2405.06219","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-05-10T03:06:24Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"39811b73fd799209cb062a7537adb2fcafab16398dfae6aa0527af38254e57c9","abstract_canon_sha256":"aab9e3a46ba3aca4b57c5a15347f12f3d443e051d89295454a07c5d5944b6148"},"schema_version":"1.0"},"canonical_sha256":"8ceb4280bdad3aa07d9415e2a10534aa6d16664706b4ea78d3921e57ae3977ab","source":{"kind":"arxiv","id":"2405.06219","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.06219","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"arxiv_version","alias_value":"2405.06219v3","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.06219","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_12","alias_value":"RTVUFAF5VU5K","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_16","alias_value":"RTVUFAF5VU5KA7MU","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_8","alias_value":"RTVUFAF5","created_at":"2026-07-05T09:34:13Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:RTVUFAF5VU5KA7MUCXRKCBJUVJ","target":"record","payload":{"canonical_record":{"source":{"id":"2405.06219","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-05-10T03:06:24Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"39811b73fd799209cb062a7537adb2fcafab16398dfae6aa0527af38254e57c9","abstract_canon_sha256":"aab9e3a46ba3aca4b57c5a15347f12f3d443e051d89295454a07c5d5944b6148"},"schema_version":"1.0"},"canonical_sha256":"8ceb4280bdad3aa07d9415e2a10534aa6d16664706b4ea78d3921e57ae3977ab","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:34:13.616088Z","signature_b64":"T+rADw47leVCZOHajwrevbbLCHaqa03W45pajDiXSBN9FrsY1o8O4NGTVPxmu1jJui1uWs5uZQXwzcj+2XmFBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8ceb4280bdad3aa07d9415e2a10534aa6d16664706b4ea78d3921e57ae3977ab","last_reissued_at":"2026-07-05T09:34:13.615451Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:34:13.615451Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2405.06219","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:34:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FIVp/dOf5akJBi8kB6cAhiAUmSRESLidjt02WEeVBh5SxbkPw6+9VrC0YaXnUC/+oDlV4aGHOmhupOPB/XO9BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T22:16:34.872580Z"},"content_sha256":"f744adb5c91febff1bf3b9a5abf55d565ac1547a6c4265f22aa0174dbacaa1bf","schema_version":"1.0","event_id":"sha256:f744adb5c91febff1bf3b9a5abf55d565ac1547a6c4265f22aa0174dbacaa1bf"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:RTVUFAF5VU5KA7MUCXRKCBJUVJ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"SKVQ: Sliding-window Key and Value Cache Quantization for Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Dahua Lin, Haojie Duanmu, Jiangfei Duan, Xingcheng Zhang, Xiuhong Li, Zhihang Yuan","submitted_at":"2024-05-10T03:06:24Z","abstract_excerpt":"Large language models (LLMs) can now handle longer sequences of tokens, enabling complex tasks like book understanding and generating lengthy novels. However, the key-value (KV) cache required for LLMs consumes substantial memory as context length increasing, becoming the bottleneck for deployment. In this paper, we present a strategy called SKVQ, which stands for sliding-window KV cache quantization, to address the issue of extremely low bitwidth KV cache quantization. To achieve this, SKVQ rearranges the channels of the KV cache in order to improve the similarity of channels in quantization "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.06219","kind":"arxiv","version":3},"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/2405.06219/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:34:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/hehSu/ErEF0XiARkN9hQfPZZisjgCW1R8k7QO60R4VivXVqOvBCE/I0/9A5GyreDev70OPdHyuKrkBbjZX/BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T22:16:34.873599Z"},"content_sha256":"94f68c786264562ac17e15dfa8a5b3970082a94be7826151614f506622ec28ad","schema_version":"1.0","event_id":"sha256:94f68c786264562ac17e15dfa8a5b3970082a94be7826151614f506622ec28ad"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RTVUFAF5VU5KA7MUCXRKCBJUVJ/bundle.json","state_url":"https://pith.science/pith/RTVUFAF5VU5KA7MUCXRKCBJUVJ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RTVUFAF5VU5KA7MUCXRKCBJUVJ/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-03T22:16:34Z","links":{"resolver":"https://pith.science/pith/RTVUFAF5VU5KA7MUCXRKCBJUVJ","bundle":"https://pith.science/pith/RTVUFAF5VU5KA7MUCXRKCBJUVJ/bundle.json","state":"https://pith.science/pith/RTVUFAF5VU5KA7MUCXRKCBJUVJ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RTVUFAF5VU5KA7MUCXRKCBJUVJ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:RTVUFAF5VU5KA7MUCXRKCBJUVJ","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"aab9e3a46ba3aca4b57c5a15347f12f3d443e051d89295454a07c5d5944b6148","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-05-10T03:06:24Z","title_canon_sha256":"39811b73fd799209cb062a7537adb2fcafab16398dfae6aa0527af38254e57c9"},"schema_version":"1.0","source":{"id":"2405.06219","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.06219","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"arxiv_version","alias_value":"2405.06219v3","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.06219","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_12","alias_value":"RTVUFAF5VU5K","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_16","alias_value":"RTVUFAF5VU5KA7MU","created_at":"2026-07-05T09:34:13Z"},{"alias_kind":"pith_short_8","alias_value":"RTVUFAF5","created_at":"2026-07-05T09:34:13Z"}],"graph_snapshots":[{"event_id":"sha256:94f68c786264562ac17e15dfa8a5b3970082a94be7826151614f506622ec28ad","target":"graph","created_at":"2026-07-05T09:34:13Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2405.06219/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) can now handle longer sequences of tokens, enabling complex tasks like book understanding and generating lengthy novels. However, the key-value (KV) cache required for LLMs consumes substantial memory as context length increasing, becoming the bottleneck for deployment. In this paper, we present a strategy called SKVQ, which stands for sliding-window KV cache quantization, to address the issue of extremely low bitwidth KV cache quantization. To achieve this, SKVQ rearranges the channels of the KV cache in order to improve the similarity of channels in quantization ","authors_text":"Dahua Lin, Haojie Duanmu, Jiangfei Duan, Xingcheng Zhang, Xiuhong Li, Zhihang Yuan","cross_cats":["cs.CL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-05-10T03:06:24Z","title":"SKVQ: Sliding-window Key and Value Cache Quantization for Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.06219","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:f744adb5c91febff1bf3b9a5abf55d565ac1547a6c4265f22aa0174dbacaa1bf","target":"record","created_at":"2026-07-05T09:34:13Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"aab9e3a46ba3aca4b57c5a15347f12f3d443e051d89295454a07c5d5944b6148","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-05-10T03:06:24Z","title_canon_sha256":"39811b73fd799209cb062a7537adb2fcafab16398dfae6aa0527af38254e57c9"},"schema_version":"1.0","source":{"id":"2405.06219","kind":"arxiv","version":3}},"canonical_sha256":"8ceb4280bdad3aa07d9415e2a10534aa6d16664706b4ea78d3921e57ae3977ab","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8ceb4280bdad3aa07d9415e2a10534aa6d16664706b4ea78d3921e57ae3977ab","first_computed_at":"2026-07-05T09:34:13.615451Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:34:13.615451Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"T+rADw47leVCZOHajwrevbbLCHaqa03W45pajDiXSBN9FrsY1o8O4NGTVPxmu1jJui1uWs5uZQXwzcj+2XmFBA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:34:13.616088Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.06219","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f744adb5c91febff1bf3b9a5abf55d565ac1547a6c4265f22aa0174dbacaa1bf","sha256:94f68c786264562ac17e15dfa8a5b3970082a94be7826151614f506622ec28ad"],"state_sha256":"8cc9f9b34b3795a5c4f2b2e1dd3c485d8d2040a195cf4ade03b4639de1b37f02"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/OwWT1M684m3njXxd1JnHn0jDFPln0VmXRwn7kqK+zVrYuQNGwxLrHq1H/8GKzBrdaIMuEvxF0J7M+Eb4B3qCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-03T22:16:34.880647Z","bundle_sha256":"9f71d476fe8d8c23967241ca7c671a253f2b9f0afb9d97e59f40407d8ce0a6f8"}}