{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:3OU3AZ7GHOPP6DMIAR5CKXO3K6","short_pith_number":"pith:3OU3AZ7G","canonical_record":{"source":{"id":"2503.23294","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T03:20:34Z","cross_cats_sorted":[],"title_canon_sha256":"cae934ce80414b902e7088e6a51512a4b62a6a35a48799c3744f2a030cbc875b","abstract_canon_sha256":"42ff105610df7c855ef04aa8dd728dddd8c44d47bcd26cf660b0dceed63075a1"},"schema_version":"1.0"},"canonical_sha256":"dba9b067e63b9eff0d88047a255ddb57b5efdb8305b11fc9c9413680e6bc23e7","source":{"kind":"arxiv","id":"2503.23294","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.23294","created_at":"2026-07-05T10:41:33Z"},{"alias_kind":"arxiv_version","alias_value":"2503.23294v1","created_at":"2026-07-05T10:41:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.23294","created_at":"2026-07-05T10:41:33Z"},{"alias_kind":"pith_short_12","alias_value":"3OU3AZ7GHOPP","created_at":"2026-07-05T10:41:33Z"},{"alias_kind":"pith_short_16","alias_value":"3OU3AZ7GHOPP6DMI","created_at":"2026-07-05T10:41:33Z"},{"alias_kind":"pith_short_8","alias_value":"3OU3AZ7G","created_at":"2026-07-05T10:41:33Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:3OU3AZ7GHOPP6DMIAR5CKXO3K6","target":"record","payload":{"canonical_record":{"source":{"id":"2503.23294","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T03:20:34Z","cross_cats_sorted":[],"title_canon_sha256":"cae934ce80414b902e7088e6a51512a4b62a6a35a48799c3744f2a030cbc875b","abstract_canon_sha256":"42ff105610df7c855ef04aa8dd728dddd8c44d47bcd26cf660b0dceed63075a1"},"schema_version":"1.0"},"canonical_sha256":"dba9b067e63b9eff0d88047a255ddb57b5efdb8305b11fc9c9413680e6bc23e7","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:41:33.899713Z","signature_b64":"H+YonMp0M6T3nPabIM6nga0EmzUui24PwLQ587o1Q+zvXtRq1+19FayiVR7QzjevDOhgr3Nh/mNqXqamh25QBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dba9b067e63b9eff0d88047a255ddb57b5efdb8305b11fc9c9413680e6bc23e7","last_reissued_at":"2026-07-05T10:41:33.899229Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:41:33.899229Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.23294","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:41:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VxU6ilct27D8SQsaCY3mIrjMjN8eXzSLfgGF9ZAyz4hqh/4qscl83aPgNCW7cxlfmmnRQr/2HSSeOoMeLEvoBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T01:04:37.810841Z"},"content_sha256":"94ebcaf1edc87096055f0f10abd0b7667ed796e6b2877ec92dc4995d98b4006d","schema_version":"1.0","event_id":"sha256:94ebcaf1edc87096055f0f10abd0b7667ed796e6b2877ec92dc4995d98b4006d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:3OU3AZ7GHOPP6DMIAR5CKXO3K6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Cocktail: Chunk-Adaptive Mixed-Precision Quantization for Long-Context LLM Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bin Zhang, Jianzong Wang, Jiguang Wan, Wei Tao, Xiaoyang Qu","submitted_at":"2025-03-30T03:20:34Z","abstract_excerpt":"Recently, large language models (LLMs) have been able to handle longer and longer contexts. However, a context that is too long may cause intolerant inference latency and GPU memory usage. Existing methods propose mixed-precision quantization to the key-value (KV) cache in LLMs based on token granularity, which is time-consuming in the search process and hardware inefficient during computation. This paper introduces a novel approach called Cocktail, which employs chunk-adaptive mixed-precision quantization to optimize the KV cache. Cocktail consists of two modules: chunk-level quantization sea"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.23294","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/2503.23294/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":"2026-07-05T10:41:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"pSXaVVuvnrtmg0NgFUE5fCG/MDqJBStoBQAL8o3tb+5nd/8Ar39Suszpklz9O4IVODGJe9JZhUqZo2CjeCF7DA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T01:04:37.811335Z"},"content_sha256":"ecf6ba2a8379c8c2fb9df3b956ac34c10ef231087e20bdbe96f712e9fac61d88","schema_version":"1.0","event_id":"sha256:ecf6ba2a8379c8c2fb9df3b956ac34c10ef231087e20bdbe96f712e9fac61d88"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3OU3AZ7GHOPP6DMIAR5CKXO3K6/bundle.json","state_url":"https://pith.science/pith/3OU3AZ7GHOPP6DMIAR5CKXO3K6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3OU3AZ7GHOPP6DMIAR5CKXO3K6/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":"2026-08-04T01:04:37Z","links":{"resolver":"https://pith.science/pith/3OU3AZ7GHOPP6DMIAR5CKXO3K6","bundle":"https://pith.science/pith/3OU3AZ7GHOPP6DMIAR5CKXO3K6/bundle.json","state":"https://pith.science/pith/3OU3AZ7GHOPP6DMIAR5CKXO3K6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3OU3AZ7GHOPP6DMIAR5CKXO3K6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:3OU3AZ7GHOPP6DMIAR5CKXO3K6","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":"42ff105610df7c855ef04aa8dd728dddd8c44d47bcd26cf660b0dceed63075a1","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T03:20:34Z","title_canon_sha256":"cae934ce80414b902e7088e6a51512a4b62a6a35a48799c3744f2a030cbc875b"},"schema_version":"1.0","source":{"id":"2503.23294","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.23294","created_at":"2026-07-05T10:41:33Z"},{"alias_kind":"arxiv_version","alias_value":"2503.23294v1","created_at":"2026-07-05T10:41:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.23294","created_at":"2026-07-05T10:41:33Z"},{"alias_kind":"pith_short_12","alias_value":"3OU3AZ7GHOPP","created_at":"2026-07-05T10:41:33Z"},{"alias_kind":"pith_short_16","alias_value":"3OU3AZ7GHOPP6DMI","created_at":"2026-07-05T10:41:33Z"},{"alias_kind":"pith_short_8","alias_value":"3OU3AZ7G","created_at":"2026-07-05T10:41:33Z"}],"graph_snapshots":[{"event_id":"sha256:ecf6ba2a8379c8c2fb9df3b956ac34c10ef231087e20bdbe96f712e9fac61d88","target":"graph","created_at":"2026-07-05T10:41:33Z","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/2503.23294/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recently, large language models (LLMs) have been able to handle longer and longer contexts. However, a context that is too long may cause intolerant inference latency and GPU memory usage. Existing methods propose mixed-precision quantization to the key-value (KV) cache in LLMs based on token granularity, which is time-consuming in the search process and hardware inefficient during computation. This paper introduces a novel approach called Cocktail, which employs chunk-adaptive mixed-precision quantization to optimize the KV cache. Cocktail consists of two modules: chunk-level quantization sea","authors_text":"Bin Zhang, Jianzong Wang, Jiguang Wan, Wei Tao, Xiaoyang Qu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T03:20:34Z","title":"Cocktail: Chunk-Adaptive Mixed-Precision Quantization for Long-Context LLM Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.23294","kind":"arxiv","version":1},"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":"sha256:94ebcaf1edc87096055f0f10abd0b7667ed796e6b2877ec92dc4995d98b4006d","target":"record","created_at":"2026-07-05T10:41:33Z","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":"42ff105610df7c855ef04aa8dd728dddd8c44d47bcd26cf660b0dceed63075a1","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T03:20:34Z","title_canon_sha256":"cae934ce80414b902e7088e6a51512a4b62a6a35a48799c3744f2a030cbc875b"},"schema_version":"1.0","source":{"id":"2503.23294","kind":"arxiv","version":1}},"canonical_sha256":"dba9b067e63b9eff0d88047a255ddb57b5efdb8305b11fc9c9413680e6bc23e7","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dba9b067e63b9eff0d88047a255ddb57b5efdb8305b11fc9c9413680e6bc23e7","first_computed_at":"2026-07-05T10:41:33.899229Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:41:33.899229Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"H+YonMp0M6T3nPabIM6nga0EmzUui24PwLQ587o1Q+zvXtRq1+19FayiVR7QzjevDOhgr3Nh/mNqXqamh25QBw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:41:33.899713Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.23294","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:94ebcaf1edc87096055f0f10abd0b7667ed796e6b2877ec92dc4995d98b4006d","sha256:ecf6ba2a8379c8c2fb9df3b956ac34c10ef231087e20bdbe96f712e9fac61d88"],"state_sha256":"e5365fb566a21bbd1508a881e56570414966f0a8cd1505894f19ca959dbf8db1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Qcj6xJD+VlBhZaYn2EVxvf6MvWv+MFrU1e21zGR7Pkbqbum6q5kk53Hr8Fi/1VjiQpXWULgwK4yUnxr2n5ikBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T01:04:37.815312Z","bundle_sha256":"98d13a60dab13423f9003bb6a7b7d6b4fdb712bf10964e29c24a9e97baf4fe6c"}}