{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:RZOZJWGL2OJTV4H4DHDEVPIRT2","short_pith_number":"pith:RZOZJWGL","canonical_record":{"source":{"id":"2412.05693","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-07T16:41:54Z","cross_cats_sorted":[],"title_canon_sha256":"c482a9b85b762a01bb34892d38520a2d4783251fae36841d5aaf10fc121d0611","abstract_canon_sha256":"c0e40d429a00ff1a31668663cc6b68212c462981a31d56b37401c4fd23d1d75e"},"schema_version":"1.0"},"canonical_sha256":"8e5d94d8cbd3933af0fc19c64abd119eb51a2c1f4a20dd70e19fa3dce00d9df2","source":{"kind":"arxiv","id":"2412.05693","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.05693","created_at":"2026-07-05T11:31:25Z"},{"alias_kind":"arxiv_version","alias_value":"2412.05693v3","created_at":"2026-07-05T11:31:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.05693","created_at":"2026-07-05T11:31:25Z"},{"alias_kind":"pith_short_12","alias_value":"RZOZJWGL2OJT","created_at":"2026-07-05T11:31:25Z"},{"alias_kind":"pith_short_16","alias_value":"RZOZJWGL2OJTV4H4","created_at":"2026-07-05T11:31:25Z"},{"alias_kind":"pith_short_8","alias_value":"RZOZJWGL","created_at":"2026-07-05T11:31:25Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:RZOZJWGL2OJTV4H4DHDEVPIRT2","target":"record","payload":{"canonical_record":{"source":{"id":"2412.05693","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-07T16:41:54Z","cross_cats_sorted":[],"title_canon_sha256":"c482a9b85b762a01bb34892d38520a2d4783251fae36841d5aaf10fc121d0611","abstract_canon_sha256":"c0e40d429a00ff1a31668663cc6b68212c462981a31d56b37401c4fd23d1d75e"},"schema_version":"1.0"},"canonical_sha256":"8e5d94d8cbd3933af0fc19c64abd119eb51a2c1f4a20dd70e19fa3dce00d9df2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:31:25.916271Z","signature_b64":"aDC7BeKLqfcU++7Bvu2CFqN6URTGAJphmOk2VlhVudbm3T5oXNRgNP9VelRdJ+sVmKY3ai/1cNJb+5D4g3UADA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8e5d94d8cbd3933af0fc19c64abd119eb51a2c1f4a20dd70e19fa3dce00d9df2","last_reissued_at":"2026-07-05T11:31:25.915890Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:31:25.915890Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.05693","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":"2026-07-05T11:31:25Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5dS3fzg0eOlA+sP2TadKswhpSYZBikDy1oeHiiRlpqBeann+KBUH5DxHYBzU48+iZQpzpU7yxvv4v0BcEFvFCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T06:03:47.892384Z"},"content_sha256":"734916ffa9cfc1fe4b3a5c1888191e9436971cf3a42c677d0a659a4f79006b84","schema_version":"1.0","event_id":"sha256:734916ffa9cfc1fe4b3a5c1888191e9436971cf3a42c677d0a659a4f79006b84"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:RZOZJWGL2OJTV4H4DHDEVPIRT2","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Batch-Max: Higher LLM Throughput using Larger Batch Sizes and KV Cache Compression","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Boxing Chen, Mehdi Rezagholizadeh, Michael R. Metel","submitted_at":"2024-12-07T16:41:54Z","abstract_excerpt":"Several works have developed eviction policies to remove key-value (KV) pairs from the KV cache for more efficient inference. The focus has been on compressing the KV cache after the input prompt has been processed for faster token generation. In settings with limited GPU memory, and when the input context is longer than the generation length, we show that by also compressing the KV cache during the input processing phase, larger batch sizes can be used resulting in significantly higher throughput while still maintaining the original model's accuracy."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.05693","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/2412.05693/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":"2026-07-05T11:31:25Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fvNBEnGDnAykHXiyML5qE8a8g4j9/eIRSJCPzc9YytGHx5R+Sr1cyRlhKIDUWxEjeJVJ0Nn2rrvt08HuMafkDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T06:03:47.892919Z"},"content_sha256":"7b94c623cd0d87dca4f54aae282085b68e8d528e7cfcfaeb7575aa5b384d965a","schema_version":"1.0","event_id":"sha256:7b94c623cd0d87dca4f54aae282085b68e8d528e7cfcfaeb7575aa5b384d965a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RZOZJWGL2OJTV4H4DHDEVPIRT2/bundle.json","state_url":"https://pith.science/pith/RZOZJWGL2OJTV4H4DHDEVPIRT2/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RZOZJWGL2OJTV4H4DHDEVPIRT2/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":"2026-08-18T06:03:47Z","links":{"resolver":"https://pith.science/pith/RZOZJWGL2OJTV4H4DHDEVPIRT2","bundle":"https://pith.science/pith/RZOZJWGL2OJTV4H4DHDEVPIRT2/bundle.json","state":"https://pith.science/pith/RZOZJWGL2OJTV4H4DHDEVPIRT2/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RZOZJWGL2OJTV4H4DHDEVPIRT2/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:RZOZJWGL2OJTV4H4DHDEVPIRT2","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":"c0e40d429a00ff1a31668663cc6b68212c462981a31d56b37401c4fd23d1d75e","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-07T16:41:54Z","title_canon_sha256":"c482a9b85b762a01bb34892d38520a2d4783251fae36841d5aaf10fc121d0611"},"schema_version":"1.0","source":{"id":"2412.05693","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.05693","created_at":"2026-07-05T11:31:25Z"},{"alias_kind":"arxiv_version","alias_value":"2412.05693v3","created_at":"2026-07-05T11:31:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.05693","created_at":"2026-07-05T11:31:25Z"},{"alias_kind":"pith_short_12","alias_value":"RZOZJWGL2OJT","created_at":"2026-07-05T11:31:25Z"},{"alias_kind":"pith_short_16","alias_value":"RZOZJWGL2OJTV4H4","created_at":"2026-07-05T11:31:25Z"},{"alias_kind":"pith_short_8","alias_value":"RZOZJWGL","created_at":"2026-07-05T11:31:25Z"}],"graph_snapshots":[{"event_id":"sha256:7b94c623cd0d87dca4f54aae282085b68e8d528e7cfcfaeb7575aa5b384d965a","target":"graph","created_at":"2026-07-05T11:31:25Z","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/2412.05693/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Several works have developed eviction policies to remove key-value (KV) pairs from the KV cache for more efficient inference. The focus has been on compressing the KV cache after the input prompt has been processed for faster token generation. In settings with limited GPU memory, and when the input context is longer than the generation length, we show that by also compressing the KV cache during the input processing phase, larger batch sizes can be used resulting in significantly higher throughput while still maintaining the original model's accuracy.","authors_text":"Boxing Chen, Mehdi Rezagholizadeh, Michael R. Metel","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-07T16:41:54Z","title":"Batch-Max: Higher LLM Throughput using Larger Batch Sizes and KV Cache Compression"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.05693","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":"sha256:734916ffa9cfc1fe4b3a5c1888191e9436971cf3a42c677d0a659a4f79006b84","target":"record","created_at":"2026-07-05T11:31:25Z","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":"c0e40d429a00ff1a31668663cc6b68212c462981a31d56b37401c4fd23d1d75e","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-07T16:41:54Z","title_canon_sha256":"c482a9b85b762a01bb34892d38520a2d4783251fae36841d5aaf10fc121d0611"},"schema_version":"1.0","source":{"id":"2412.05693","kind":"arxiv","version":3}},"canonical_sha256":"8e5d94d8cbd3933af0fc19c64abd119eb51a2c1f4a20dd70e19fa3dce00d9df2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8e5d94d8cbd3933af0fc19c64abd119eb51a2c1f4a20dd70e19fa3dce00d9df2","first_computed_at":"2026-07-05T11:31:25.915890Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:31:25.915890Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"aDC7BeKLqfcU++7Bvu2CFqN6URTGAJphmOk2VlhVudbm3T5oXNRgNP9VelRdJ+sVmKY3ai/1cNJb+5D4g3UADA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:31:25.916271Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.05693","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:734916ffa9cfc1fe4b3a5c1888191e9436971cf3a42c677d0a659a4f79006b84","sha256:7b94c623cd0d87dca4f54aae282085b68e8d528e7cfcfaeb7575aa5b384d965a"],"state_sha256":"c0a7be7268b33d53c119031a7b722d86514e171143c553c134f8a2f5516d3955"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Tz5fds7pUGQmgy+UNwm4gJLrkxLdzf7GVs5bShQA/PfuJ7IkvQCy3Ev7esWnbsnjiFydtbQz5DttYzZGVRmwDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-18T06:03:47.899911Z","bundle_sha256":"94296b1deb3c9100e8dc7afcdc002512b7a89ee6b2811b8e2164441bedab5102"}}