{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:NNY4K32IRAHFCRTQ3WNS2DXZS6","short_pith_number":"pith:NNY4K32I","canonical_record":{"source":{"id":"2503.08879","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-11T20:45:02Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"2f5a1cf4eacd092032391bce1a5a633144b13a4e0c094b5940eae2946f1a41ec","abstract_canon_sha256":"9d5a422b3bcb403b42906da5e9fcf878285847d939da1098260b3682ce9ef0ff"},"schema_version":"1.0"},"canonical_sha256":"6b71c56f48880e514670dd9b2d0ef997be8c5f084732b7d493c65b71b12187d1","source":{"kind":"arxiv","id":"2503.08879","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.08879","created_at":"2026-07-05T10:29:43Z"},{"alias_kind":"arxiv_version","alias_value":"2503.08879v1","created_at":"2026-07-05T10:29:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.08879","created_at":"2026-07-05T10:29:43Z"},{"alias_kind":"pith_short_12","alias_value":"NNY4K32IRAHF","created_at":"2026-07-05T10:29:43Z"},{"alias_kind":"pith_short_16","alias_value":"NNY4K32IRAHFCRTQ","created_at":"2026-07-05T10:29:43Z"},{"alias_kind":"pith_short_8","alias_value":"NNY4K32I","created_at":"2026-07-05T10:29:43Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:NNY4K32IRAHFCRTQ3WNS2DXZS6","target":"record","payload":{"canonical_record":{"source":{"id":"2503.08879","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-11T20:45:02Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"2f5a1cf4eacd092032391bce1a5a633144b13a4e0c094b5940eae2946f1a41ec","abstract_canon_sha256":"9d5a422b3bcb403b42906da5e9fcf878285847d939da1098260b3682ce9ef0ff"},"schema_version":"1.0"},"canonical_sha256":"6b71c56f48880e514670dd9b2d0ef997be8c5f084732b7d493c65b71b12187d1","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:29:43.069411Z","signature_b64":"uEbnTafmdq9jWx7XxoTmt+DBVMVSZsBp0yr0OZPfrnoSWhYu8SQz0iZ8LIiRf6XOnOOmfuVp68NlM759NT0PCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6b71c56f48880e514670dd9b2d0ef997be8c5f084732b7d493c65b71b12187d1","last_reissued_at":"2026-07-05T10:29:43.068888Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:29:43.068888Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.08879","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":"2026-07-05T10:29:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HQzl7bpPQRkN0RIbxMtCS/hdLyOQQiALfshJzIOqGwi8L2Mq/PQWnWe4/Pn9TqqnoIQ5mTEP2JkxJFzuhtEeAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T03:56:15.851012Z"},"content_sha256":"e0cab3d8cc169124c976e65d6d3b2443c4ac745c4d4e9ae54dcb8e86a89bff0c","schema_version":"1.0","event_id":"sha256:e0cab3d8cc169124c976e65d6d3b2443c4ac745c4d4e9ae54dcb8e86a89bff0c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:NNY4K32IRAHFCRTQ3WNS2DXZS6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LLMs Know What to Drop: Self-Attention Guided KV Cache Eviction for Efficient Long-Context Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Bo Li, Changran Hu, Chen Wu, Darshan Gandhi, Guangtao Wang, Jonathan Li, Shubhangi Upasani, Urmish Thakker","submitted_at":"2025-03-11T20:45:02Z","abstract_excerpt":"Efficient long-context inference is critical as large language models (LLMs) adopt context windows of ranging from 128K to 1M tokens. However, the growing key-value (KV) cache and the high computational complexity of attention create significant bottlenecks in memory usage and latency. In this paper, we find that attention in diverse long-context tasks exhibits sparsity, and LLMs implicitly \"know\" which tokens can be dropped or evicted at the head level after the pre-filling stage. Based on this insight, we propose Self-Attention Guided Eviction~(SAGE-KV), a simple and effective KV eviction ca"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.08879","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/2503.08879/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":"2026-07-05T10:29:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"++vX37NVs1W6bbZ+Xxfc/n14+OuTfVsBFs8YyZostNca4c04kHav3HhAZqYiemuSmSVJdvggIOloZIU7rIsoAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T03:56:15.851408Z"},"content_sha256":"572a3612cea9e30c0192c7f927aedbc112fe21c439e1bc04206be710fab148b1","schema_version":"1.0","event_id":"sha256:572a3612cea9e30c0192c7f927aedbc112fe21c439e1bc04206be710fab148b1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NNY4K32IRAHFCRTQ3WNS2DXZS6/bundle.json","state_url":"https://pith.science/pith/NNY4K32IRAHFCRTQ3WNS2DXZS6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NNY4K32IRAHFCRTQ3WNS2DXZS6/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":"2026-07-07T03:56:15Z","links":{"resolver":"https://pith.science/pith/NNY4K32IRAHFCRTQ3WNS2DXZS6","bundle":"https://pith.science/pith/NNY4K32IRAHFCRTQ3WNS2DXZS6/bundle.json","state":"https://pith.science/pith/NNY4K32IRAHFCRTQ3WNS2DXZS6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NNY4K32IRAHFCRTQ3WNS2DXZS6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:NNY4K32IRAHFCRTQ3WNS2DXZS6","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":"9d5a422b3bcb403b42906da5e9fcf878285847d939da1098260b3682ce9ef0ff","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-11T20:45:02Z","title_canon_sha256":"2f5a1cf4eacd092032391bce1a5a633144b13a4e0c094b5940eae2946f1a41ec"},"schema_version":"1.0","source":{"id":"2503.08879","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.08879","created_at":"2026-07-05T10:29:43Z"},{"alias_kind":"arxiv_version","alias_value":"2503.08879v1","created_at":"2026-07-05T10:29:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.08879","created_at":"2026-07-05T10:29:43Z"},{"alias_kind":"pith_short_12","alias_value":"NNY4K32IRAHF","created_at":"2026-07-05T10:29:43Z"},{"alias_kind":"pith_short_16","alias_value":"NNY4K32IRAHFCRTQ","created_at":"2026-07-05T10:29:43Z"},{"alias_kind":"pith_short_8","alias_value":"NNY4K32I","created_at":"2026-07-05T10:29:43Z"}],"graph_snapshots":[{"event_id":"sha256:572a3612cea9e30c0192c7f927aedbc112fe21c439e1bc04206be710fab148b1","target":"graph","created_at":"2026-07-05T10:29:43Z","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/2503.08879/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Efficient long-context inference is critical as large language models (LLMs) adopt context windows of ranging from 128K to 1M tokens. However, the growing key-value (KV) cache and the high computational complexity of attention create significant bottlenecks in memory usage and latency. In this paper, we find that attention in diverse long-context tasks exhibits sparsity, and LLMs implicitly \"know\" which tokens can be dropped or evicted at the head level after the pre-filling stage. Based on this insight, we propose Self-Attention Guided Eviction~(SAGE-KV), a simple and effective KV eviction ca","authors_text":"Bo Li, Changran Hu, Chen Wu, Darshan Gandhi, Guangtao Wang, Jonathan Li, Shubhangi Upasani, Urmish Thakker","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-11T20:45:02Z","title":"LLMs Know What to Drop: Self-Attention Guided KV Cache Eviction for Efficient Long-Context Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.08879","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:e0cab3d8cc169124c976e65d6d3b2443c4ac745c4d4e9ae54dcb8e86a89bff0c","target":"record","created_at":"2026-07-05T10:29:43Z","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":"9d5a422b3bcb403b42906da5e9fcf878285847d939da1098260b3682ce9ef0ff","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-11T20:45:02Z","title_canon_sha256":"2f5a1cf4eacd092032391bce1a5a633144b13a4e0c094b5940eae2946f1a41ec"},"schema_version":"1.0","source":{"id":"2503.08879","kind":"arxiv","version":1}},"canonical_sha256":"6b71c56f48880e514670dd9b2d0ef997be8c5f084732b7d493c65b71b12187d1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6b71c56f48880e514670dd9b2d0ef997be8c5f084732b7d493c65b71b12187d1","first_computed_at":"2026-07-05T10:29:43.068888Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:29:43.068888Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"uEbnTafmdq9jWx7XxoTmt+DBVMVSZsBp0yr0OZPfrnoSWhYu8SQz0iZ8LIiRf6XOnOOmfuVp68NlM759NT0PCw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:29:43.069411Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.08879","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e0cab3d8cc169124c976e65d6d3b2443c4ac745c4d4e9ae54dcb8e86a89bff0c","sha256:572a3612cea9e30c0192c7f927aedbc112fe21c439e1bc04206be710fab148b1"],"state_sha256":"9ab05a63a11d74e583c0377bec01b340f7a667037d9fa4d294fdcf346fc88e6a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"syqiL0drrkCgiWZ0YXW9lHyzyv6d6t/kl1j2T4AlXSWMXwisn2Gz66gfiodH3BpD7KS/AL26Dis81QdiOzwBAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T03:56:15.853553Z","bundle_sha256":"745a1e1f87fe1c4396b4efff24453dfa9da4acdddab1b82b7799588fe9ad349b"}}