{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:H47YDJN6QLTUEOMDTYUCMZAPG6","short_pith_number":"pith:H47YDJN6","schema_version":"1.0","canonical_sha256":"3f3f81a5be82e74239839e2826640f37b9ef78fcb07785b0d2ae2fea5bfe1685","source":{"kind":"arxiv","id":"2604.23878","version":3},"attestation_state":"computed","paper":{"title":"ZenBrain: A Neuroscience-Inspired 7-Layer Memory Architecture for Autonomous AI Systems","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A 7-layer neuroscience-inspired memory system for AI reaches 91 percent of long-context oracle accuracy at 1/106th the token cost.","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Alexander Bering","submitted_at":"2026-04-26T20:39:19Z","abstract_excerpt":"ZenBrain is a seven-layer, neuroscience-derived memory architecture for LLM agents that unifies fifteen mechanisms - from Two-Factor synaptic consolidation to a Simulation-Selection sleep loop - under a single MemoryCoordinator: nine foundational algorithms plus six Predictive Memory Architecture components. No system among those we survey integrates more than two of them. Ablating each mechanism separately exposes an effect we call cooperative masking. Under moderate load, fourteen of the fifteen ablations look costless - the architecture reads as mostly dead weight. Raising decay to 0.25/day"},"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":"2604.23878","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-04-26T20:39:19Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"c2f94e3057661e312d8f27ecbf417ec5149b318e4c2d9ce5d9c91890b750c5fd","abstract_canon_sha256":"64a15dd32a35ce611e8255ddf4f18a94360c851b3b45b0f8c5e09b80b1997692"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T01:20:42.480979Z","signature_b64":"hMu/9Gjw0ZXh+sysonK8KyMR4DbHPXd/aT7Q3OrjaAjPbDWAS1MB/bhOI3GAKfvPqskNENcoSxjM49AyKw0gDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3f3f81a5be82e74239839e2826640f37b9ef78fcb07785b0d2ae2fea5bfe1685","last_reissued_at":"2026-08-11T01:20:42.478004Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T01:20:42.478004Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ZenBrain: A Neuroscience-Inspired 7-Layer Memory Architecture for Autonomous AI Systems","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A 7-layer neuroscience-inspired memory system for AI reaches 91 percent of long-context oracle accuracy at 1/106th the token cost.","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Alexander Bering","submitted_at":"2026-04-26T20:39:19Z","abstract_excerpt":"ZenBrain is a seven-layer, neuroscience-derived memory architecture for LLM agents that unifies fifteen mechanisms - from Two-Factor synaptic consolidation to a Simulation-Selection sleep loop - under a single MemoryCoordinator: nine foundational algorithms plus six Predictive Memory Architecture components. No system among those we survey integrates more than two of them. Ablating each mechanism separately exposes an effect we call cooperative masking. Under moderate load, fourteen of the fifteen ablations look costless - the architecture reads as mostly dead weight. Raising decay to 0.25/day"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"On LongMemEval-500, ZenBrain matches a long-context oracle's binary-judge accuracy to within 4.5 pp (47.7% vs. 52.2%; 91.3%) at 1/106th of the per-query token cost, and wins all 12 head-to-head answer-quality cells against Letta, Mem0, and A-Mem under Bonferroni correction.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the 15 neuroscience mechanisms translate directly into effective AI components without introducing hidden interactions or benchmark-specific artifacts, and that the 60-day stress ablations with 10 seeds fully isolate each mechanism's contribution.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ZenBrain unifies 15 neuroscience mechanisms into a 7-layer memory system that achieves near-oracle long-context accuracy at 1/106th token cost and outperforms prior memory architectures in controlled comparisons.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A 7-layer neuroscience-inspired memory system for AI reaches 91 percent of long-context oracle accuracy at 1/106th the token cost.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"ca65bb5b91c8162456e8b8023660656cd9ce8e5063d62bcef771c6bda870f30b"},"source":{"id":"2604.23878","kind":"arxiv","version":3},"verdict":{"id":"6518284c-133e-4c5b-b763-f10b0ca81917","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T06:06:09.910847Z","strongest_claim":"On LongMemEval-500, ZenBrain matches a long-context oracle's binary-judge accuracy to within 4.5 pp (47.7% vs. 52.2%; 91.3%) at 1/106th of the per-query token cost, and wins all 12 head-to-head answer-quality cells against Letta, Mem0, and A-Mem under Bonferroni correction.","one_line_summary":"ZenBrain unifies 15 neuroscience mechanisms into a 7-layer memory system that achieves near-oracle long-context accuracy at 1/106th token cost and outperforms prior memory architectures in controlled comparisons.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the 15 neuroscience mechanisms translate directly into effective AI components without introducing hidden interactions or benchmark-specific artifacts, and that the 60-day stress ablations with 10 seeds fully isolate each mechanism's contribution.","pith_extraction_headline":"A 7-layer neuroscience-inspired memory system for AI reaches 91 percent of long-context oracle accuracy at 1/106th the token cost."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.23878/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T07:43:01.924504Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T22:42:14.560350Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"5936f54ac55623f7b392bcf48a34d52f90639226d27bbca51a713bb29de8bb43"},"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":"2604.23878","created_at":"2026-08-11T01:20:42.478723+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.23878v3","created_at":"2026-08-11T01:20:42.478723+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.23878","created_at":"2026-08-11T01:20:42.478723+00:00"},{"alias_kind":"pith_short_12","alias_value":"H47YDJN6QLTU","created_at":"2026-08-11T01:20:42.478723+00:00"},{"alias_kind":"pith_short_16","alias_value":"H47YDJN6QLTUEOMD","created_at":"2026-08-11T01:20:42.478723+00:00"},{"alias_kind":"pith_short_8","alias_value":"H47YDJN6","created_at":"2026-08-11T01:20:42.478723+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.12428","citing_title":"MindMemOS: A Portable and Self-Evolving Memory Operating Layer for AI Agents","ref_index":4,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/H47YDJN6QLTUEOMDTYUCMZAPG6","json":"https://pith.science/pith/H47YDJN6QLTUEOMDTYUCMZAPG6.json","graph_json":"https://pith.science/api/pith-number/H47YDJN6QLTUEOMDTYUCMZAPG6/graph.json","events_json":"https://pith.science/api/pith-number/H47YDJN6QLTUEOMDTYUCMZAPG6/events.json","paper":"https://pith.science/paper/H47YDJN6"},"agent_actions":{"view_html":"https://pith.science/pith/H47YDJN6QLTUEOMDTYUCMZAPG6","download_json":"https://pith.science/pith/H47YDJN6QLTUEOMDTYUCMZAPG6.json","view_paper":"https://pith.science/paper/H47YDJN6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.23878&json=true","fetch_graph":"https://pith.science/api/pith-number/H47YDJN6QLTUEOMDTYUCMZAPG6/graph.json","fetch_events":"https://pith.science/api/pith-number/H47YDJN6QLTUEOMDTYUCMZAPG6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H47YDJN6QLTUEOMDTYUCMZAPG6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H47YDJN6QLTUEOMDTYUCMZAPG6/action/storage_attestation","attest_author":"https://pith.science/pith/H47YDJN6QLTUEOMDTYUCMZAPG6/action/author_attestation","sign_citation":"https://pith.science/pith/H47YDJN6QLTUEOMDTYUCMZAPG6/action/citation_signature","submit_replication":"https://pith.science/pith/H47YDJN6QLTUEOMDTYUCMZAPG6/action/replication_record"}},"created_at":"2026-08-11T01:20:42.478723+00:00","updated_at":"2026-08-11T01:20:42.478723+00:00"}