{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:KTFTDCNAAGH4IKKMGHSKTQOJS4","short_pith_number":"pith:KTFTDCNA","schema_version":"1.0","canonical_sha256":"54cb3189a0018fc4294c31e4a9c1c99730253fa439e3b0fe6266b2ed09002785","source":{"kind":"arxiv","id":"2509.11155","version":1},"attestation_state":"computed","paper":{"title":"AQUA: Attention via QUery mAgnitudes for Memory and Compute Efficient Inference in LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Balaraman Ravindran, Santhosh G S, Saurav Prakash","submitted_at":"2025-09-14T08:20:48Z","abstract_excerpt":"The quadratic complexity of the attention mechanism remains a fundamental barrier to scaling Large Language Models (LLMs) to longer contexts, creating a critical bottleneck in both computation and memory. To address this, we introduce AQUA (Attention via QUery mAgnitudes) a novel and versatile approximation strategy that significantly reduces the cost of attention with a graceful performance trade-off. Our method operates in two phases: an efficient offline step where we compute a universal, language agnostic projection matrix via SVD on a calibration dataset, and an online inference step wher"},"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":"2509.11155","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-09-14T08:20:48Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"eff9abd70c339319af47486ee3c74df83714ff31d68e28ad9276c1b9eea3ff62","abstract_canon_sha256":"b98fa9adbd55c2659facb826c59afa55d0217e46305878d88685be440b16b1e2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:11:34.512746Z","signature_b64":"HO8wpdKXkTgvGTC6XYR7FS+gZjlH3XF6Rwgynxi92WDQCBnhnVyHjLp2t/Px8f0MkHGkoHbEwvBoufNDZi7ZAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"54cb3189a0018fc4294c31e4a9c1c99730253fa439e3b0fe6266b2ed09002785","last_reissued_at":"2026-07-05T12:11:34.512237Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:11:34.512237Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AQUA: Attention via QUery mAgnitudes for Memory and Compute Efficient Inference in LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Balaraman Ravindran, Santhosh G S, Saurav Prakash","submitted_at":"2025-09-14T08:20:48Z","abstract_excerpt":"The quadratic complexity of the attention mechanism remains a fundamental barrier to scaling Large Language Models (LLMs) to longer contexts, creating a critical bottleneck in both computation and memory. To address this, we introduce AQUA (Attention via QUery mAgnitudes) a novel and versatile approximation strategy that significantly reduces the cost of attention with a graceful performance trade-off. Our method operates in two phases: an efficient offline step where we compute a universal, language agnostic projection matrix via SVD on a calibration dataset, and an online inference step wher"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.11155","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/2509.11155/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":"2509.11155","created_at":"2026-07-05T12:11:34.512292+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.11155v1","created_at":"2026-07-05T12:11:34.512292+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.11155","created_at":"2026-07-05T12:11:34.512292+00:00"},{"alias_kind":"pith_short_12","alias_value":"KTFTDCNAAGH4","created_at":"2026-07-05T12:11:34.512292+00:00"},{"alias_kind":"pith_short_16","alias_value":"KTFTDCNAAGH4IKKM","created_at":"2026-07-05T12:11:34.512292+00:00"},{"alias_kind":"pith_short_8","alias_value":"KTFTDCNA","created_at":"2026-07-05T12:11:34.512292+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/KTFTDCNAAGH4IKKMGHSKTQOJS4","json":"https://pith.science/pith/KTFTDCNAAGH4IKKMGHSKTQOJS4.json","graph_json":"https://pith.science/api/pith-number/KTFTDCNAAGH4IKKMGHSKTQOJS4/graph.json","events_json":"https://pith.science/api/pith-number/KTFTDCNAAGH4IKKMGHSKTQOJS4/events.json","paper":"https://pith.science/paper/KTFTDCNA"},"agent_actions":{"view_html":"https://pith.science/pith/KTFTDCNAAGH4IKKMGHSKTQOJS4","download_json":"https://pith.science/pith/KTFTDCNAAGH4IKKMGHSKTQOJS4.json","view_paper":"https://pith.science/paper/KTFTDCNA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.11155&json=true","fetch_graph":"https://pith.science/api/pith-number/KTFTDCNAAGH4IKKMGHSKTQOJS4/graph.json","fetch_events":"https://pith.science/api/pith-number/KTFTDCNAAGH4IKKMGHSKTQOJS4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KTFTDCNAAGH4IKKMGHSKTQOJS4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KTFTDCNAAGH4IKKMGHSKTQOJS4/action/storage_attestation","attest_author":"https://pith.science/pith/KTFTDCNAAGH4IKKMGHSKTQOJS4/action/author_attestation","sign_citation":"https://pith.science/pith/KTFTDCNAAGH4IKKMGHSKTQOJS4/action/citation_signature","submit_replication":"https://pith.science/pith/KTFTDCNAAGH4IKKMGHSKTQOJS4/action/replication_record"}},"created_at":"2026-07-05T12:11:34.512292+00:00","updated_at":"2026-07-05T12:11:34.512292+00:00"}