{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:I5I3UHUE2SI5D2FELGOS6CMOJD","short_pith_number":"pith:I5I3UHUE","schema_version":"1.0","canonical_sha256":"4751ba1e84d491d1e8a4599d2f098e48ef1d4cc5332dd4422cdbe45866f06785","source":{"kind":"arxiv","id":"2607.11614","version":1},"attestation_state":"computed","paper":{"title":"Extending LLM Context via Associative Recurrent Memory","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Artem Shelmanov, Aydar Bulatov, Gleb Kuzmin, Ilia Sochenkov, Ivan Rodkin, Lyudmila Rvanova, Mikhail Burtsev, Mikhail Katkov, Misha Tsodyks, Timothy Baldwin, Yuri Kuratov","submitted_at":"2026-07-13T14:37:24Z","abstract_excerpt":"Extending the context length of large language models (LLMs) is critical for many real-world applications, yet standard transformers remain constrained by quadratic compute and linear memory scaling. In this work, we investigate the Associative Recurrent Memory Transformer (ARMT) as a practical approach for enabling long-context processing in LLMs, constant memory scaling, and better efficiency. We make three main contributions. First, we construct two domain-specific long-context datasets designed to evaluate realistic workloads, focusing on narrow-domain fine-tuning scenarios. Second, we pro"},"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":"2607.11614","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-13T14:37:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d087672539be881d82b04693b434782193312c9a0bc0e07055f8c30ee5c4d506","abstract_canon_sha256":"42744a27ba310e638012ac4633d6005364c50ada5abd92de8821f2dd3a79f268"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-14T02:22:15.038038Z","signature_b64":"T41LT6BZWukG4rEip46Ccq0Z0b78UQmVd869hXJvqJKfggMihzZGuFAuZQupN8sh9L3BmEwf3vGXXNROe0I/Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4751ba1e84d491d1e8a4599d2f098e48ef1d4cc5332dd4422cdbe45866f06785","last_reissued_at":"2026-07-14T02:22:15.037165Z","signature_status":"signed_v1","first_computed_at":"2026-07-14T02:22:15.037165Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Extending LLM Context via Associative Recurrent Memory","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Artem Shelmanov, Aydar Bulatov, Gleb Kuzmin, Ilia Sochenkov, Ivan Rodkin, Lyudmila Rvanova, Mikhail Burtsev, Mikhail Katkov, Misha Tsodyks, Timothy Baldwin, Yuri Kuratov","submitted_at":"2026-07-13T14:37:24Z","abstract_excerpt":"Extending the context length of large language models (LLMs) is critical for many real-world applications, yet standard transformers remain constrained by quadratic compute and linear memory scaling. In this work, we investigate the Associative Recurrent Memory Transformer (ARMT) as a practical approach for enabling long-context processing in LLMs, constant memory scaling, and better efficiency. We make three main contributions. First, we construct two domain-specific long-context datasets designed to evaluate realistic workloads, focusing on narrow-domain fine-tuning scenarios. Second, we pro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.11614","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/2607.11614/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":"2607.11614","created_at":"2026-07-14T02:22:15.037624+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.11614v1","created_at":"2026-07-14T02:22:15.037624+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.11614","created_at":"2026-07-14T02:22:15.037624+00:00"},{"alias_kind":"pith_short_12","alias_value":"I5I3UHUE2SI5","created_at":"2026-07-14T02:22:15.037624+00:00"},{"alias_kind":"pith_short_16","alias_value":"I5I3UHUE2SI5D2FE","created_at":"2026-07-14T02:22:15.037624+00:00"},{"alias_kind":"pith_short_8","alias_value":"I5I3UHUE","created_at":"2026-07-14T02:22:15.037624+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/I5I3UHUE2SI5D2FELGOS6CMOJD","json":"https://pith.science/pith/I5I3UHUE2SI5D2FELGOS6CMOJD.json","graph_json":"https://pith.science/api/pith-number/I5I3UHUE2SI5D2FELGOS6CMOJD/graph.json","events_json":"https://pith.science/api/pith-number/I5I3UHUE2SI5D2FELGOS6CMOJD/events.json","paper":"https://pith.science/paper/I5I3UHUE"},"agent_actions":{"view_html":"https://pith.science/pith/I5I3UHUE2SI5D2FELGOS6CMOJD","download_json":"https://pith.science/pith/I5I3UHUE2SI5D2FELGOS6CMOJD.json","view_paper":"https://pith.science/paper/I5I3UHUE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.11614&json=true","fetch_graph":"https://pith.science/api/pith-number/I5I3UHUE2SI5D2FELGOS6CMOJD/graph.json","fetch_events":"https://pith.science/api/pith-number/I5I3UHUE2SI5D2FELGOS6CMOJD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/I5I3UHUE2SI5D2FELGOS6CMOJD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/I5I3UHUE2SI5D2FELGOS6CMOJD/action/storage_attestation","attest_author":"https://pith.science/pith/I5I3UHUE2SI5D2FELGOS6CMOJD/action/author_attestation","sign_citation":"https://pith.science/pith/I5I3UHUE2SI5D2FELGOS6CMOJD/action/citation_signature","submit_replication":"https://pith.science/pith/I5I3UHUE2SI5D2FELGOS6CMOJD/action/replication_record"}},"created_at":"2026-07-14T02:22:15.037624+00:00","updated_at":"2026-07-14T02:22:15.037624+00:00"}