{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:A2A2JL3M5JEH6K7UW2EV4MJAHK","short_pith_number":"pith:A2A2JL3M","schema_version":"1.0","canonical_sha256":"0681a4af6cea487f2bf4b6895e31203aade9d2a31d2d3709599548ec1d22d6a1","source":{"kind":"arxiv","id":"2403.17312","version":1},"attestation_state":"computed","paper":{"title":"ALISA: Accelerating Large Language Model Inference via Sparsity-Aware KV Caching","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.PF"],"primary_cat":"cs.AI","authors_text":"Di Wu, Jun Wang, Youpeng Zhao","submitted_at":"2024-03-26T01:46:34Z","abstract_excerpt":"The Transformer architecture has significantly advanced natural language processing (NLP) and has been foundational in developing large language models (LLMs) such as LLaMA and OPT, which have come to dominate a broad range of NLP tasks. Despite their superior accuracy, LLMs present unique challenges in practical inference, concerning the compute and memory-intensive nature. Thanks to the autoregressive characteristic of LLM inference, KV caching for the attention layers in Transformers can effectively accelerate LLM inference by substituting quadratic-complexity computation with linear-comple"},"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":"2403.17312","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-03-26T01:46:34Z","cross_cats_sorted":["cs.LG","cs.PF"],"title_canon_sha256":"d318cd6c2c6ad871d356563196da6d13c02a1eb6cca75c2e2724a75cef9348a8","abstract_canon_sha256":"937677504d0ecfe93952cb5ec39e94a4bb2af99f5d9a437a55d974f5da77b5af"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:00:48.760890Z","signature_b64":"hckAXKNW0x26g/s1BF9PuwD9e6cm1izTA2WQ8PdaxM9k1gw4YJvIVaxQxLjqCnDRT2zKx+cVyseAPTdXzQcHCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0681a4af6cea487f2bf4b6895e31203aade9d2a31d2d3709599548ec1d22d6a1","last_reissued_at":"2026-07-05T08:00:48.760469Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:00:48.760469Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ALISA: Accelerating Large Language Model Inference via Sparsity-Aware KV Caching","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.PF"],"primary_cat":"cs.AI","authors_text":"Di Wu, Jun Wang, Youpeng Zhao","submitted_at":"2024-03-26T01:46:34Z","abstract_excerpt":"The Transformer architecture has significantly advanced natural language processing (NLP) and has been foundational in developing large language models (LLMs) such as LLaMA and OPT, which have come to dominate a broad range of NLP tasks. Despite their superior accuracy, LLMs present unique challenges in practical inference, concerning the compute and memory-intensive nature. Thanks to the autoregressive characteristic of LLM inference, KV caching for the attention layers in Transformers can effectively accelerate LLM inference by substituting quadratic-complexity computation with linear-comple"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.17312","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/2403.17312/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":"2403.17312","created_at":"2026-07-05T08:00:48.760526+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.17312v1","created_at":"2026-07-05T08:00:48.760526+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.17312","created_at":"2026-07-05T08:00:48.760526+00:00"},{"alias_kind":"pith_short_12","alias_value":"A2A2JL3M5JEH","created_at":"2026-07-05T08:00:48.760526+00:00"},{"alias_kind":"pith_short_16","alias_value":"A2A2JL3M5JEH6K7U","created_at":"2026-07-05T08:00:48.760526+00:00"},{"alias_kind":"pith_short_8","alias_value":"A2A2JL3M","created_at":"2026-07-05T08:00:48.760526+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.20179","citing_title":"TIDE: Efficient and Lossless MoE Diffusion LLM Inference with I/O-aware Expert Offload","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00831","citing_title":"GhostServe: A Lightweight Checkpointing System in the Shadow for Fault-Tolerant LLM Serving","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/A2A2JL3M5JEH6K7UW2EV4MJAHK","json":"https://pith.science/pith/A2A2JL3M5JEH6K7UW2EV4MJAHK.json","graph_json":"https://pith.science/api/pith-number/A2A2JL3M5JEH6K7UW2EV4MJAHK/graph.json","events_json":"https://pith.science/api/pith-number/A2A2JL3M5JEH6K7UW2EV4MJAHK/events.json","paper":"https://pith.science/paper/A2A2JL3M"},"agent_actions":{"view_html":"https://pith.science/pith/A2A2JL3M5JEH6K7UW2EV4MJAHK","download_json":"https://pith.science/pith/A2A2JL3M5JEH6K7UW2EV4MJAHK.json","view_paper":"https://pith.science/paper/A2A2JL3M","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.17312&json=true","fetch_graph":"https://pith.science/api/pith-number/A2A2JL3M5JEH6K7UW2EV4MJAHK/graph.json","fetch_events":"https://pith.science/api/pith-number/A2A2JL3M5JEH6K7UW2EV4MJAHK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/A2A2JL3M5JEH6K7UW2EV4MJAHK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/A2A2JL3M5JEH6K7UW2EV4MJAHK/action/storage_attestation","attest_author":"https://pith.science/pith/A2A2JL3M5JEH6K7UW2EV4MJAHK/action/author_attestation","sign_citation":"https://pith.science/pith/A2A2JL3M5JEH6K7UW2EV4MJAHK/action/citation_signature","submit_replication":"https://pith.science/pith/A2A2JL3M5JEH6K7UW2EV4MJAHK/action/replication_record"}},"created_at":"2026-07-05T08:00:48.760526+00:00","updated_at":"2026-07-05T08:00:48.760526+00:00"}