{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:C4W7KMFQ64K4DJNPZXGRNT67FM","short_pith_number":"pith:C4W7KMFQ","schema_version":"1.0","canonical_sha256":"172df530b0f715c1a5afcdcd16cfdf2b0128f82fafba7363e3d5516728e68770","source":{"kind":"arxiv","id":"2407.04841","version":2},"attestation_state":"computed","paper":{"title":"Associative Recurrent Memory Transformer","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Aydar Bulatov, Ivan Rodkin, Mikhail Burtsev, Yuri Kuratov","submitted_at":"2024-07-05T19:57:49Z","abstract_excerpt":"This paper addresses the challenge of creating a neural architecture for very long sequences that requires constant time for processing new information at each time step. Our approach, Associative Recurrent Memory Transformer (ARMT), is based on transformer self-attention for local context and segment-level recurrence for storage of task specific information distributed over a long context. We demonstrate that ARMT outperfors existing alternatives in associative retrieval tasks and sets a new performance record in the recent BABILong multi-task long-context benchmark by answering single-fact q"},"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":"2407.04841","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-05T19:57:49Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"1508ec2b448ee1625dbfca31e0cda2f50df7a3c084699fd4d46059f639cd4ec7","abstract_canon_sha256":"32af3abf00f461cb7a5ab23a96c72dabda33ac116974e42f1e6803929916f9ee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:14:04.273138Z","signature_b64":"Da2wO8pvr/yrTDM0eSPpIWKnCI6RC8dBntPkV/zoJkTgxF+/fEYWSdd//fi7eFZzP1MvvABKz5PROon5n1agCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"172df530b0f715c1a5afcdcd16cfdf2b0128f82fafba7363e3d5516728e68770","last_reissued_at":"2026-07-05T10:14:04.272646Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:14:04.272646Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Associative Recurrent Memory Transformer","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Aydar Bulatov, Ivan Rodkin, Mikhail Burtsev, Yuri Kuratov","submitted_at":"2024-07-05T19:57:49Z","abstract_excerpt":"This paper addresses the challenge of creating a neural architecture for very long sequences that requires constant time for processing new information at each time step. Our approach, Associative Recurrent Memory Transformer (ARMT), is based on transformer self-attention for local context and segment-level recurrence for storage of task specific information distributed over a long context. We demonstrate that ARMT outperfors existing alternatives in associative retrieval tasks and sets a new performance record in the recent BABILong multi-task long-context benchmark by answering single-fact q"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.04841","kind":"arxiv","version":2},"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/2407.04841/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":"2407.04841","created_at":"2026-07-05T10:14:04.272704+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.04841v2","created_at":"2026-07-05T10:14:04.272704+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.04841","created_at":"2026-07-05T10:14:04.272704+00:00"},{"alias_kind":"pith_short_12","alias_value":"C4W7KMFQ64K4","created_at":"2026-07-05T10:14:04.272704+00:00"},{"alias_kind":"pith_short_16","alias_value":"C4W7KMFQ64K4DJNP","created_at":"2026-07-05T10:14:04.272704+00:00"},{"alias_kind":"pith_short_8","alias_value":"C4W7KMFQ","created_at":"2026-07-05T10:14:04.272704+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.03979","citing_title":"Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories","ref_index":87,"is_internal_anchor":false},{"citing_arxiv_id":"2508.16745","citing_title":"Beyond Memorization: Extending Reasoning Depth with Recurrence, Memory and Test-Time Compute Scaling","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2511.07328","citing_title":"Q-RAG: Long Context Multi-step Retrieval via Value-based Embedder Training","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2501.00663","citing_title":"Titans: Learning to Memorize at Test Time","ref_index":90,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/C4W7KMFQ64K4DJNPZXGRNT67FM","json":"https://pith.science/pith/C4W7KMFQ64K4DJNPZXGRNT67FM.json","graph_json":"https://pith.science/api/pith-number/C4W7KMFQ64K4DJNPZXGRNT67FM/graph.json","events_json":"https://pith.science/api/pith-number/C4W7KMFQ64K4DJNPZXGRNT67FM/events.json","paper":"https://pith.science/paper/C4W7KMFQ"},"agent_actions":{"view_html":"https://pith.science/pith/C4W7KMFQ64K4DJNPZXGRNT67FM","download_json":"https://pith.science/pith/C4W7KMFQ64K4DJNPZXGRNT67FM.json","view_paper":"https://pith.science/paper/C4W7KMFQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.04841&json=true","fetch_graph":"https://pith.science/api/pith-number/C4W7KMFQ64K4DJNPZXGRNT67FM/graph.json","fetch_events":"https://pith.science/api/pith-number/C4W7KMFQ64K4DJNPZXGRNT67FM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/C4W7KMFQ64K4DJNPZXGRNT67FM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/C4W7KMFQ64K4DJNPZXGRNT67FM/action/storage_attestation","attest_author":"https://pith.science/pith/C4W7KMFQ64K4DJNPZXGRNT67FM/action/author_attestation","sign_citation":"https://pith.science/pith/C4W7KMFQ64K4DJNPZXGRNT67FM/action/citation_signature","submit_replication":"https://pith.science/pith/C4W7KMFQ64K4DJNPZXGRNT67FM/action/replication_record"}},"created_at":"2026-07-05T10:14:04.272704+00:00","updated_at":"2026-07-05T10:14:04.272704+00:00"}