{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:MTIBYTUEHUJVBD7GXIY253PDXV","short_pith_number":"pith:MTIBYTUE","canonical_record":{"source":{"id":"2608.03555","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2026-08-04T12:25:45Z","cross_cats_sorted":[],"title_canon_sha256":"a2b1bad57ac8a0d93af42f2e21245fa27ce86a77bc6340ff6359b69364ede935","abstract_canon_sha256":"782018a6ab022033ab662a9b587c1f499c573ae8329dfafca1ff16beb05f53bc"},"schema_version":"1.0"},"canonical_sha256":"64d01c4e843d13508fe6ba31aeede3bd5559a5454d37ec4117e5c774b797537f","source":{"kind":"arxiv","id":"2608.03555","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.03555","created_at":"2026-08-05T01:36:38Z"},{"alias_kind":"arxiv_version","alias_value":"2608.03555v1","created_at":"2026-08-05T01:36:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.03555","created_at":"2026-08-05T01:36:38Z"},{"alias_kind":"pith_short_12","alias_value":"MTIBYTUEHUJV","created_at":"2026-08-05T01:36:38Z"},{"alias_kind":"pith_short_16","alias_value":"MTIBYTUEHUJVBD7G","created_at":"2026-08-05T01:36:38Z"},{"alias_kind":"pith_short_8","alias_value":"MTIBYTUE","created_at":"2026-08-05T01:36:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:MTIBYTUEHUJVBD7GXIY253PDXV","target":"record","payload":{"canonical_record":{"source":{"id":"2608.03555","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2026-08-04T12:25:45Z","cross_cats_sorted":[],"title_canon_sha256":"a2b1bad57ac8a0d93af42f2e21245fa27ce86a77bc6340ff6359b69364ede935","abstract_canon_sha256":"782018a6ab022033ab662a9b587c1f499c573ae8329dfafca1ff16beb05f53bc"},"schema_version":"1.0"},"canonical_sha256":"64d01c4e843d13508fe6ba31aeede3bd5559a5454d37ec4117e5c774b797537f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-05T01:36:38.376317Z","signature_b64":"5Fj08LfFZI6dWJchUlBOphIx+/x50z43azNbbrJwxFjGwPKMa06dJ7DagTnDBc3OtENs+y7rTdcYQMklDZz/DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"64d01c4e843d13508fe6ba31aeede3bd5559a5454d37ec4117e5c774b797537f","last_reissued_at":"2026-08-05T01:36:38.374743Z","signature_status":"signed_v1","first_computed_at":"2026-08-05T01:36:38.374743Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.03555","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":"2026-08-05T01:36:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5gk0gkV0MsxApf19WAFyZvVzQvsvDHSv9EdN2V5RrBL+rhuDy7m+flF7B+V9dWwdkcDcBLWflU/Pb/W8b8jGCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-22T09:55:53.816970Z"},"content_sha256":"e01710b047bf68eb09d076564d93f0496de42f07582b9e0939f67e674e2c9337","schema_version":"1.0","event_id":"sha256:e01710b047bf68eb09d076564d93f0496de42f07582b9e0939f67e674e2c9337"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:MTIBYTUEHUJVBD7GXIY253PDXV","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Heterogeneous LLM Serving with General-Purpose Processing-Near-Memory for Retrieval-Based Sparse Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Gwangsun Kim, Hyungkyu Ham, Junhyeong Bae, Myeongjae Jeon, Seungheon Lee","submitted_at":"2026-08-04T12:25:45Z","abstract_excerpt":"This paper presents a heterogeneous decode-phase serving system that relocates the KV cache out of GPU memory, motivated by the retrieval-based sparse attention that recent frontier LLMs adopt to serve million-token contexts. It partitions a decode step by operation type: GPU nodes hold the model weights and execute the projections and MoE layers, while processing-near-memory (PNM) nodes hold the KV cache and index keys and execute every operation that reads them. We first show that the assumptions behind prior PIM and PNM designs no longer hold for these operations, and derive four design req"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.03555","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/2608.03555/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":"2026-08-05T01:36:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"O329YqbkDedWXNtrVfjx58t6gFrpg+yKdTWA4LDLTRURsaxVoLMRm7Ax1SWdaXJcXxN/3MH2evD621ca23LDCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-22T09:55:53.817527Z"},"content_sha256":"446695178a4ed394984428a32f03b75e233009aef788052bb4ce4ef05e204767","schema_version":"1.0","event_id":"sha256:446695178a4ed394984428a32f03b75e233009aef788052bb4ce4ef05e204767"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MTIBYTUEHUJVBD7GXIY253PDXV/bundle.json","state_url":"https://pith.science/pith/MTIBYTUEHUJVBD7GXIY253PDXV/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MTIBYTUEHUJVBD7GXIY253PDXV/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":"2026-08-22T09:55:53Z","links":{"resolver":"https://pith.science/pith/MTIBYTUEHUJVBD7GXIY253PDXV","bundle":"https://pith.science/pith/MTIBYTUEHUJVBD7GXIY253PDXV/bundle.json","state":"https://pith.science/pith/MTIBYTUEHUJVBD7GXIY253PDXV/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MTIBYTUEHUJVBD7GXIY253PDXV/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:MTIBYTUEHUJVBD7GXIY253PDXV","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":"782018a6ab022033ab662a9b587c1f499c573ae8329dfafca1ff16beb05f53bc","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2026-08-04T12:25:45Z","title_canon_sha256":"a2b1bad57ac8a0d93af42f2e21245fa27ce86a77bc6340ff6359b69364ede935"},"schema_version":"1.0","source":{"id":"2608.03555","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.03555","created_at":"2026-08-05T01:36:38Z"},{"alias_kind":"arxiv_version","alias_value":"2608.03555v1","created_at":"2026-08-05T01:36:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.03555","created_at":"2026-08-05T01:36:38Z"},{"alias_kind":"pith_short_12","alias_value":"MTIBYTUEHUJV","created_at":"2026-08-05T01:36:38Z"},{"alias_kind":"pith_short_16","alias_value":"MTIBYTUEHUJVBD7G","created_at":"2026-08-05T01:36:38Z"},{"alias_kind":"pith_short_8","alias_value":"MTIBYTUE","created_at":"2026-08-05T01:36:38Z"}],"graph_snapshots":[{"event_id":"sha256:446695178a4ed394984428a32f03b75e233009aef788052bb4ce4ef05e204767","target":"graph","created_at":"2026-08-05T01:36:38Z","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/2608.03555/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper presents a heterogeneous decode-phase serving system that relocates the KV cache out of GPU memory, motivated by the retrieval-based sparse attention that recent frontier LLMs adopt to serve million-token contexts. It partitions a decode step by operation type: GPU nodes hold the model weights and execute the projections and MoE layers, while processing-near-memory (PNM) nodes hold the KV cache and index keys and execute every operation that reads them. We first show that the assumptions behind prior PIM and PNM designs no longer hold for these operations, and derive four design req","authors_text":"Gwangsun Kim, Hyungkyu Ham, Junhyeong Bae, Myeongjae Jeon, Seungheon Lee","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2026-08-04T12:25:45Z","title":"Heterogeneous LLM Serving with General-Purpose Processing-Near-Memory for Retrieval-Based Sparse Attention"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.03555","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:e01710b047bf68eb09d076564d93f0496de42f07582b9e0939f67e674e2c9337","target":"record","created_at":"2026-08-05T01:36:38Z","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":"782018a6ab022033ab662a9b587c1f499c573ae8329dfafca1ff16beb05f53bc","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2026-08-04T12:25:45Z","title_canon_sha256":"a2b1bad57ac8a0d93af42f2e21245fa27ce86a77bc6340ff6359b69364ede935"},"schema_version":"1.0","source":{"id":"2608.03555","kind":"arxiv","version":1}},"canonical_sha256":"64d01c4e843d13508fe6ba31aeede3bd5559a5454d37ec4117e5c774b797537f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"64d01c4e843d13508fe6ba31aeede3bd5559a5454d37ec4117e5c774b797537f","first_computed_at":"2026-08-05T01:36:38.374743Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-05T01:36:38.374743Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5Fj08LfFZI6dWJchUlBOphIx+/x50z43azNbbrJwxFjGwPKMa06dJ7DagTnDBc3OtENs+y7rTdcYQMklDZz/DA==","signature_status":"signed_v1","signed_at":"2026-08-05T01:36:38.376317Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.03555","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e01710b047bf68eb09d076564d93f0496de42f07582b9e0939f67e674e2c9337","sha256:446695178a4ed394984428a32f03b75e233009aef788052bb4ce4ef05e204767"],"state_sha256":"b3825ae39f96425e2747b3d771e8e940fd4d258ce4dfb12fc0bd079c0afca6bf"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FD2WyNmgnrFMs3CIdOQgVj43I4MkXxkgweFvwbfUmMISkRJxTlv6IvsF4NE3NjJAWIOyVakv/R1NHcTLHcXLBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-22T09:55:53.822057Z","bundle_sha256":"9f822650889255be355b946b2f95cd4a0556401fe49797fad3fe3a91ccc40202"}}