{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:QVK5EKOPIZIHMFIS7ZE4GITX5O","short_pith_number":"pith:QVK5EKOP","schema_version":"1.0","canonical_sha256":"8555d229cf4650761512fe49c32277eb9736070fe8aa2f7c626747e4b6c41eec","source":{"kind":"arxiv","id":"2607.19368","version":1},"attestation_state":"computed","paper":{"title":"Spectral-LSH: Sub-Quadratic Prompt Compression via Krylov-Projected Locality-Sensitive Hashing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Ali Mahdavi, Amirfarhad Farhadi, Azaseh Zamanifar, Omid Kashefi","submitted_at":"2026-06-12T01:11:17Z","abstract_excerpt":"Long-prompt inference remains expensive because prefill attention scales quadratically with sequence length. We propose Spectral-LSH, a training-free prompt compression method that operates before the prompt enters the language model. Spectral-LSH approximates the dominant components of an implicit attention-kernel operator using a Krylov subspace method together with random features, avoiding explicit $O(N^2)$ attention-kernel materialization. It then applies SimHash in the resulting attention eigenspace to group similar tokens and aggregate them into macro-tokens with causal positional assig"},"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":"2607.19368","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-12T01:11:17Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"2167690307be7bcee0bc85e21da3a6997b6bb1782ee9df3db9f14f2f89d85a0b","abstract_canon_sha256":"658a6d5c20ceed63b28502229feee501ee2026af5c230d6162971366578f0e12"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-23T00:23:45.178878Z","signature_b64":"Rwx1vJq6NeavtRBcec0CVBChlOkMB6F7eDcM+3+d5raIv856LEmYxKmFLFY/8WwLX+GSabYPLrSHb+AkSyTlAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8555d229cf4650761512fe49c32277eb9736070fe8aa2f7c626747e4b6c41eec","last_reissued_at":"2026-07-23T00:23:45.177970Z","signature_status":"signed_v1","first_computed_at":"2026-07-23T00:23:45.177970Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Spectral-LSH: Sub-Quadratic Prompt Compression via Krylov-Projected Locality-Sensitive Hashing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Ali Mahdavi, Amirfarhad Farhadi, Azaseh Zamanifar, Omid Kashefi","submitted_at":"2026-06-12T01:11:17Z","abstract_excerpt":"Long-prompt inference remains expensive because prefill attention scales quadratically with sequence length. We propose Spectral-LSH, a training-free prompt compression method that operates before the prompt enters the language model. Spectral-LSH approximates the dominant components of an implicit attention-kernel operator using a Krylov subspace method together with random features, avoiding explicit $O(N^2)$ attention-kernel materialization. It then applies SimHash in the resulting attention eigenspace to group similar tokens and aggregate them into macro-tokens with causal positional assig"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.19368","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/2607.19368/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":"2607.19368","created_at":"2026-07-23T00:23:45.178420+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.19368v1","created_at":"2026-07-23T00:23:45.178420+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.19368","created_at":"2026-07-23T00:23:45.178420+00:00"},{"alias_kind":"pith_short_12","alias_value":"QVK5EKOPIZIH","created_at":"2026-07-23T00:23:45.178420+00:00"},{"alias_kind":"pith_short_16","alias_value":"QVK5EKOPIZIHMFIS","created_at":"2026-07-23T00:23:45.178420+00:00"},{"alias_kind":"pith_short_8","alias_value":"QVK5EKOP","created_at":"2026-07-23T00:23:45.178420+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/QVK5EKOPIZIHMFIS7ZE4GITX5O","json":"https://pith.science/pith/QVK5EKOPIZIHMFIS7ZE4GITX5O.json","graph_json":"https://pith.science/api/pith-number/QVK5EKOPIZIHMFIS7ZE4GITX5O/graph.json","events_json":"https://pith.science/api/pith-number/QVK5EKOPIZIHMFIS7ZE4GITX5O/events.json","paper":"https://pith.science/paper/QVK5EKOP"},"agent_actions":{"view_html":"https://pith.science/pith/QVK5EKOPIZIHMFIS7ZE4GITX5O","download_json":"https://pith.science/pith/QVK5EKOPIZIHMFIS7ZE4GITX5O.json","view_paper":"https://pith.science/paper/QVK5EKOP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.19368&json=true","fetch_graph":"https://pith.science/api/pith-number/QVK5EKOPIZIHMFIS7ZE4GITX5O/graph.json","fetch_events":"https://pith.science/api/pith-number/QVK5EKOPIZIHMFIS7ZE4GITX5O/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QVK5EKOPIZIHMFIS7ZE4GITX5O/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QVK5EKOPIZIHMFIS7ZE4GITX5O/action/storage_attestation","attest_author":"https://pith.science/pith/QVK5EKOPIZIHMFIS7ZE4GITX5O/action/author_attestation","sign_citation":"https://pith.science/pith/QVK5EKOPIZIHMFIS7ZE4GITX5O/action/citation_signature","submit_replication":"https://pith.science/pith/QVK5EKOPIZIHMFIS7ZE4GITX5O/action/replication_record"}},"created_at":"2026-07-23T00:23:45.178420+00:00","updated_at":"2026-07-23T00:23:45.178420+00:00"}