{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:46JVBM74ZEMMPP7PMFYYP6LMPN","short_pith_number":"pith:46JVBM74","schema_version":"1.0","canonical_sha256":"e79350b3fcc918c7bfef617187f96c7b5201c576aa6f362c985d5191acdbd012","source":{"kind":"arxiv","id":"2312.11514","version":3},"attestation_state":"computed","paper":{"title":"LLM in a flash: Efficient Large Language Model Inference with Limited Memory","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Carlo C Del Mundo, Dmitry Belenko, Iman Mirzadeh, Karen Khatamifard, Keivan Alizadeh, Mehrdad Farajtabar, Minsik Cho, Mohammad Rastegari","submitted_at":"2023-12-12T18:57:08Z","abstract_excerpt":"Large language models (LLMs) are central to modern natural language processing, delivering exceptional performance in various tasks. However, their substantial computational and memory requirements present challenges, especially for devices with limited DRAM capacity. This paper tackles the challenge of efficiently running LLMs that exceed the available DRAM capacity by storing the model parameters in flash memory, but bringing them on demand to DRAM. Our method involves constructing an inference cost model that takes into account the characteristics of flash memory, guiding us to optimize in "},"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":"2312.11514","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-12-12T18:57:08Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"e6bf99ac86993acf29ba676c3385ac647f2975d0e550d5369015a9d38d9b8fb0","abstract_canon_sha256":"bb9d88d62cbdf82435c6d4da7319c19524c296e9ddeb5267dd63d44dad8a7734"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:50:26.426922Z","signature_b64":"ho1JQWwVjqwcDcLcpAoMEZaGrunj27dx84dmzJAYTYRXssoJiM8uVAEv7TQhbtGuz4asf1CfgMuO7D5qSfIODw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e79350b3fcc918c7bfef617187f96c7b5201c576aa6f362c985d5191acdbd012","last_reissued_at":"2026-07-05T08:50:26.426499Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:50:26.426499Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LLM in a flash: Efficient Large Language Model Inference with Limited Memory","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Carlo C Del Mundo, Dmitry Belenko, Iman Mirzadeh, Karen Khatamifard, Keivan Alizadeh, Mehrdad Farajtabar, Minsik Cho, Mohammad Rastegari","submitted_at":"2023-12-12T18:57:08Z","abstract_excerpt":"Large language models (LLMs) are central to modern natural language processing, delivering exceptional performance in various tasks. However, their substantial computational and memory requirements present challenges, especially for devices with limited DRAM capacity. This paper tackles the challenge of efficiently running LLMs that exceed the available DRAM capacity by storing the model parameters in flash memory, but bringing them on demand to DRAM. Our method involves constructing an inference cost model that takes into account the characteristics of flash memory, guiding us to optimize in "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.11514","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/2312.11514/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":"2312.11514","created_at":"2026-07-05T08:50:26.426557+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.11514v3","created_at":"2026-07-05T08:50:26.426557+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.11514","created_at":"2026-07-05T08:50:26.426557+00:00"},{"alias_kind":"pith_short_12","alias_value":"46JVBM74ZEMM","created_at":"2026-07-05T08:50:26.426557+00:00"},{"alias_kind":"pith_short_16","alias_value":"46JVBM74ZEMMPP7P","created_at":"2026-07-05T08:50:26.426557+00:00"},{"alias_kind":"pith_short_8","alias_value":"46JVBM74","created_at":"2026-07-05T08:50:26.426557+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.23001","citing_title":"EnerInfer: Energy-Aware On-Device LLM Inference","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21428","citing_title":"Does Mixture-of-Experts Actually Help Inference on Consumer and Edge Hardware? An Empirical Study","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2606.12556","citing_title":"ITME: Inference Tiered Memory Expansion with Disaggregated CXL-Hybrid Memories","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2401.05459","citing_title":"Personal LLM Agents: Insights and Survey about the Capability, Efficiency and Security","ref_index":275,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11128","citing_title":"Technology solutions targeting the performance of gen-AI inference in resource constrained platforms","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/46JVBM74ZEMMPP7PMFYYP6LMPN","json":"https://pith.science/pith/46JVBM74ZEMMPP7PMFYYP6LMPN.json","graph_json":"https://pith.science/api/pith-number/46JVBM74ZEMMPP7PMFYYP6LMPN/graph.json","events_json":"https://pith.science/api/pith-number/46JVBM74ZEMMPP7PMFYYP6LMPN/events.json","paper":"https://pith.science/paper/46JVBM74"},"agent_actions":{"view_html":"https://pith.science/pith/46JVBM74ZEMMPP7PMFYYP6LMPN","download_json":"https://pith.science/pith/46JVBM74ZEMMPP7PMFYYP6LMPN.json","view_paper":"https://pith.science/paper/46JVBM74","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.11514&json=true","fetch_graph":"https://pith.science/api/pith-number/46JVBM74ZEMMPP7PMFYYP6LMPN/graph.json","fetch_events":"https://pith.science/api/pith-number/46JVBM74ZEMMPP7PMFYYP6LMPN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/46JVBM74ZEMMPP7PMFYYP6LMPN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/46JVBM74ZEMMPP7PMFYYP6LMPN/action/storage_attestation","attest_author":"https://pith.science/pith/46JVBM74ZEMMPP7PMFYYP6LMPN/action/author_attestation","sign_citation":"https://pith.science/pith/46JVBM74ZEMMPP7PMFYYP6LMPN/action/citation_signature","submit_replication":"https://pith.science/pith/46JVBM74ZEMMPP7PMFYYP6LMPN/action/replication_record"}},"created_at":"2026-07-05T08:50:26.426557+00:00","updated_at":"2026-07-05T08:50:26.426557+00:00"}