{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:SIHDZTSGRB6A72WCNGAD5WDOXG","short_pith_number":"pith:SIHDZTSG","schema_version":"1.0","canonical_sha256":"920e3cce46887c0feac269803ed86eb99927cc5b699fb6d2e3bf13d4169f015e","source":{"kind":"arxiv","id":"2604.17249","version":2},"attestation_state":"computed","paper":{"title":"Bit-Flip Vulnerability of Shared KV-Cache Blocks in LLM Serving Systems","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Shared KV-cache blocks in LLM serving systems can be corrupted by bit flips, causing silent but persistent changes in responses for all requests using the same prefix.","cross_cats":["cs.AR","cs.LG"],"primary_cat":"cs.CR","authors_text":"Satoshi Matsuura, Yuji Yamamoto","submitted_at":"2026-04-19T04:31:12Z","abstract_excerpt":"Rowhammer on GPU DRAM has enabled adversarial bit flips in model weights; shared KV-cache blocks in LLM serving systems present an analogous but previously unexamined target. In vLLM's Prefix Caching, these blocks exist as a single physical copy without integrity protection. Using software fault injection under ideal bit targeting, we characterize worst-case severity and identify three properties: (1) Silent divergence - 13 of 16 BF16 bit positions produce coherent but altered outputs, indistinguishable from legitimate responses without a clean baseline. (2) Selective propagation - only reques"},"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":"2604.17249","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2026-04-19T04:31:12Z","cross_cats_sorted":["cs.AR","cs.LG"],"title_canon_sha256":"dabdce088bb81c16d1ab5f379084a62025e1f8e20b663a6faf7d0b0c8f1da46e","abstract_canon_sha256":"e3fe9f41d34acb32243f1d07b7e8f72f3b01b961facd3bec3c11c2f96140c8fd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T01:05:17.338587Z","signature_b64":"JMb3SHbpCYkbX679XJAaZGQ4ddXanMapsnmYRT5bdzTYSo7Hk/kpxUtSqId0pmJdkS9XLXUNyDuwq2/60HdxAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"920e3cce46887c0feac269803ed86eb99927cc5b699fb6d2e3bf13d4169f015e","last_reissued_at":"2026-06-09T01:05:17.337921Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T01:05:17.337921Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bit-Flip Vulnerability of Shared KV-Cache Blocks in LLM Serving Systems","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Shared KV-cache blocks in LLM serving systems can be corrupted by bit flips, causing silent but persistent changes in responses for all requests using the same prefix.","cross_cats":["cs.AR","cs.LG"],"primary_cat":"cs.CR","authors_text":"Satoshi Matsuura, Yuji Yamamoto","submitted_at":"2026-04-19T04:31:12Z","abstract_excerpt":"Rowhammer on GPU DRAM has enabled adversarial bit flips in model weights; shared KV-cache blocks in LLM serving systems present an analogous but previously unexamined target. In vLLM's Prefix Caching, these blocks exist as a single physical copy without integrity protection. Using software fault injection under ideal bit targeting, we characterize worst-case severity and identify three properties: (1) Silent divergence - 13 of 16 BF16 bit positions produce coherent but altered outputs, indistinguishable from legitimate responses without a clean baseline. (2) Selective propagation - only reques"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Shared KV-cache blocks in LLM serving systems present an analogous but previously unexamined target. Using software fault injection under ideal bit targeting, we characterize worst-case severity and identify three properties: (1) Silent divergence - 13 of 16 BF16 bit positions produce coherent but altered outputs, indistinguishable from legitimate responses without a clean baseline. (2) Selective propagation - only requests sharing the targeted prefix are affected. (3) Persistent accumulation - no temporal decay occurs, so cumulative damage grows linearly with subsequent requests.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"Software fault injection under ideal bit targeting accurately represents the effects and feasibility of real Rowhammer attacks on GPU DRAM in production LLM serving systems with shared prefix caches.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Shared KV-cache blocks in LLM serving systems are vulnerable to bit flips that cause silent, selective, and persistent output changes, distinct from weight attacks, and detectable by checksums.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Shared KV-cache blocks in LLM serving systems can be corrupted by bit flips, causing silent but persistent changes in responses for all requests using the same prefix.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"949b905ac8d38e7fdec1ab289ca3e2077c07e1af3a91193cc962a40b56f01901"},"source":{"id":"2604.17249","kind":"arxiv","version":2},"verdict":{"id":"191e996f-aabf-4556-bdb2-71c7dd7ee005","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T06:31:08.495171Z","strongest_claim":"Shared KV-cache blocks in LLM serving systems present an analogous but previously unexamined target. Using software fault injection under ideal bit targeting, we characterize worst-case severity and identify three properties: (1) Silent divergence - 13 of 16 BF16 bit positions produce coherent but altered outputs, indistinguishable from legitimate responses without a clean baseline. (2) Selective propagation - only requests sharing the targeted prefix are affected. (3) Persistent accumulation - no temporal decay occurs, so cumulative damage grows linearly with subsequent requests.","one_line_summary":"Shared KV-cache blocks in LLM serving systems are vulnerable to bit flips that cause silent, selective, and persistent output changes, distinct from weight attacks, and detectable by checksums.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"Software fault injection under ideal bit targeting accurately represents the effects and feasibility of real Rowhammer attacks on GPU DRAM in production LLM serving systems with shared prefix caches.","pith_extraction_headline":"Shared KV-cache blocks in LLM serving systems can be corrupted by bit flips, causing silent but persistent changes in responses for all requests using the same prefix."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.17249/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":"2604.17249","created_at":"2026-06-09T01:05:17.338011+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.17249v2","created_at":"2026-06-09T01:05:17.338011+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.17249","created_at":"2026-06-09T01:05:17.338011+00:00"},{"alias_kind":"pith_short_12","alias_value":"SIHDZTSGRB6A","created_at":"2026-06-09T01:05:17.338011+00:00"},{"alias_kind":"pith_short_16","alias_value":"SIHDZTSGRB6A72WC","created_at":"2026-06-09T01:05:17.338011+00:00"},{"alias_kind":"pith_short_8","alias_value":"SIHDZTSG","created_at":"2026-06-09T01:05:17.338011+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/SIHDZTSGRB6A72WCNGAD5WDOXG","json":"https://pith.science/pith/SIHDZTSGRB6A72WCNGAD5WDOXG.json","graph_json":"https://pith.science/api/pith-number/SIHDZTSGRB6A72WCNGAD5WDOXG/graph.json","events_json":"https://pith.science/api/pith-number/SIHDZTSGRB6A72WCNGAD5WDOXG/events.json","paper":"https://pith.science/paper/SIHDZTSG"},"agent_actions":{"view_html":"https://pith.science/pith/SIHDZTSGRB6A72WCNGAD5WDOXG","download_json":"https://pith.science/pith/SIHDZTSGRB6A72WCNGAD5WDOXG.json","view_paper":"https://pith.science/paper/SIHDZTSG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.17249&json=true","fetch_graph":"https://pith.science/api/pith-number/SIHDZTSGRB6A72WCNGAD5WDOXG/graph.json","fetch_events":"https://pith.science/api/pith-number/SIHDZTSGRB6A72WCNGAD5WDOXG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SIHDZTSGRB6A72WCNGAD5WDOXG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SIHDZTSGRB6A72WCNGAD5WDOXG/action/storage_attestation","attest_author":"https://pith.science/pith/SIHDZTSGRB6A72WCNGAD5WDOXG/action/author_attestation","sign_citation":"https://pith.science/pith/SIHDZTSGRB6A72WCNGAD5WDOXG/action/citation_signature","submit_replication":"https://pith.science/pith/SIHDZTSGRB6A72WCNGAD5WDOXG/action/replication_record"}},"created_at":"2026-06-09T01:05:17.338011+00:00","updated_at":"2026-06-09T01:05:17.338011+00:00"}