{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:PFNGNROUXPGBDFD3EG6RHZXYRV","short_pith_number":"pith:PFNGNROU","schema_version":"1.0","canonical_sha256":"795a66c5d4bbcc11947b21bd13e6f88d7c07f596dd1e92dcb6ede22ed05a1769","source":{"kind":"arxiv","id":"2409.19315","version":2},"attestation_state":"computed","paper":{"title":"Analog In-Memory Computing Attention Mechanism for Fast and Energy-Efficient Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.AR","cs.ET"],"primary_cat":"cs.NE","authors_text":"Chirag Sudarshan, Emre Neftci, Jan Finkbeiner, John Paul Strachan, Nathan Leroux, Paul-Philipp Manea, Sebastian Siegel","submitted_at":"2024-09-28T11:00:11Z","abstract_excerpt":"Transformer networks, driven by self-attention, are central to Large Language Models. In generative Transformers, self-attention uses cache memory to store token projections, avoiding recomputation at each time step. However, GPU-stored projections must be loaded into SRAM for each new generation step, causing latency and energy bottlenecks.\n  We present a custom self-attention in-memory computing architecture based on emerging charge-based memories called gain cells, which can be efficiently written to store new tokens during sequence generation and enable parallel analog dot-product computat"},"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":"2409.19315","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.NE","submitted_at":"2024-09-28T11:00:11Z","cross_cats_sorted":["cs.AI","cs.AR","cs.ET"],"title_canon_sha256":"05b52dd58016808bc367d5336daf8582184197bf3f3907946a632b6b6b34448d","abstract_canon_sha256":"fa1a7911cbd4cb4a9f3aaa78a0905939a476e52aab18503ddd03c9b8296ecb15"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:39:58.861740Z","signature_b64":"cx+iJhTGGFMpTz8Q1r8aoQIq+yycn/jzo9TogsGryLox8g00+QhxYrLkpefmP27raT37G5mMLrT4vbsY1LQjBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"795a66c5d4bbcc11947b21bd13e6f88d7c07f596dd1e92dcb6ede22ed05a1769","last_reissued_at":"2026-07-05T09:39:58.861266Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:39:58.861266Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Analog In-Memory Computing Attention Mechanism for Fast and Energy-Efficient Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.AR","cs.ET"],"primary_cat":"cs.NE","authors_text":"Chirag Sudarshan, Emre Neftci, Jan Finkbeiner, John Paul Strachan, Nathan Leroux, Paul-Philipp Manea, Sebastian Siegel","submitted_at":"2024-09-28T11:00:11Z","abstract_excerpt":"Transformer networks, driven by self-attention, are central to Large Language Models. In generative Transformers, self-attention uses cache memory to store token projections, avoiding recomputation at each time step. However, GPU-stored projections must be loaded into SRAM for each new generation step, causing latency and energy bottlenecks.\n  We present a custom self-attention in-memory computing architecture based on emerging charge-based memories called gain cells, which can be efficiently written to store new tokens during sequence generation and enable parallel analog dot-product computat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.19315","kind":"arxiv","version":2},"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/2409.19315/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":"2409.19315","created_at":"2026-07-05T09:39:58.861324+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.19315v2","created_at":"2026-07-05T09:39:58.861324+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.19315","created_at":"2026-07-05T09:39:58.861324+00:00"},{"alias_kind":"pith_short_12","alias_value":"PFNGNROUXPGB","created_at":"2026-07-05T09:39:58.861324+00:00"},{"alias_kind":"pith_short_16","alias_value":"PFNGNROUXPGBDFD3","created_at":"2026-07-05T09:39:58.861324+00:00"},{"alias_kind":"pith_short_8","alias_value":"PFNGNROU","created_at":"2026-07-05T09:39:58.861324+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.28208","citing_title":"FCDC: Nonvolatile Charge-Domain Attention with HZO Ferroelectric Capacitors","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28208","citing_title":"FCDC: Nonvolatile Charge-Domain Attention with HZO Ferroelectric Capacitors","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28208","citing_title":"FCDC: Nonvolatile Charge-Domain Attention with HZO Ferroelectric Capacitors","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PFNGNROUXPGBDFD3EG6RHZXYRV","json":"https://pith.science/pith/PFNGNROUXPGBDFD3EG6RHZXYRV.json","graph_json":"https://pith.science/api/pith-number/PFNGNROUXPGBDFD3EG6RHZXYRV/graph.json","events_json":"https://pith.science/api/pith-number/PFNGNROUXPGBDFD3EG6RHZXYRV/events.json","paper":"https://pith.science/paper/PFNGNROU"},"agent_actions":{"view_html":"https://pith.science/pith/PFNGNROUXPGBDFD3EG6RHZXYRV","download_json":"https://pith.science/pith/PFNGNROUXPGBDFD3EG6RHZXYRV.json","view_paper":"https://pith.science/paper/PFNGNROU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.19315&json=true","fetch_graph":"https://pith.science/api/pith-number/PFNGNROUXPGBDFD3EG6RHZXYRV/graph.json","fetch_events":"https://pith.science/api/pith-number/PFNGNROUXPGBDFD3EG6RHZXYRV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PFNGNROUXPGBDFD3EG6RHZXYRV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PFNGNROUXPGBDFD3EG6RHZXYRV/action/storage_attestation","attest_author":"https://pith.science/pith/PFNGNROUXPGBDFD3EG6RHZXYRV/action/author_attestation","sign_citation":"https://pith.science/pith/PFNGNROUXPGBDFD3EG6RHZXYRV/action/citation_signature","submit_replication":"https://pith.science/pith/PFNGNROUXPGBDFD3EG6RHZXYRV/action/replication_record"}},"created_at":"2026-07-05T09:39:58.861324+00:00","updated_at":"2026-07-05T09:39:58.861324+00:00"}