{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:QOPZW2B2YFYM2D6YOCSNSC5YUO","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"d190569143980c41c63739ea5ab93edafb302492a5bf8dd424afdb714a6040b2","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-01-12T09:54:49Z","title_canon_sha256":"f356246ba4b44007608f772ef51593fefac15a455acd4eda5c5b20e0d67de9a2"},"schema_version":"1.0","source":{"id":"2601.07372","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.07372","created_at":"2026-05-17T23:38:49Z"},{"alias_kind":"arxiv_version","alias_value":"2601.07372v1","created_at":"2026-05-17T23:38:49Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.07372","created_at":"2026-05-17T23:38:49Z"},{"alias_kind":"pith_short_12","alias_value":"QOPZW2B2YFYM","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"QOPZW2B2YFYM2D6Y","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"QOPZW2B2","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:2cfef55bceced902d01380035e847c16e8f968dfbc2445f1000386902bfc9e63","target":"graph","created_at":"2026-05-17T23:38:49Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Scaling Engram to 27B parameters achieves superior performance over a strictly iso-parameter and iso-FLOPs MoE baseline, with notable gains in reasoning (BBH +5.0, ARC-Challenge +3.7) and long-context retrieval (Multi-Query NIAH: 84.2 to 97.0)."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The U-shaped scaling law for sparsity allocation between MoE computation and Engram memory generalizes beyond the tested model sizes and tasks, and the observed mechanistic benefits (relieving early layers, freeing attention) are causally due to the memory module rather than confounding factors in the experimental setup."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Engram adds conditional memory via scalable lookup to LLMs, outperforming iso-parameter MoE baselines on reasoning and long-context tasks by following a U-shaped scaling law for allocating between computation and memory."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Engram introduces conditional memory as a new sparsity axis that lets large language models perform direct O(1) knowledge lookups instead of computing retrieval."}],"snapshot_sha256":"a145764756cd9bd88e0cffbe98c0fde785ed2b626bc47c3eb146e930678aa54f"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"52a4a16841deb854fa05fa6ad541d62379d1bb2674cd11ab0c6baf8ac188b567"},"paper":{"abstract_excerpt":"While Mixture-of-Experts (MoE) scales capacity via conditional computation, Transformers lack a native primitive for knowledge lookup, forcing them to inefficiently simulate retrieval through computation. To address this, we introduce conditional memory as a complementary sparsity axis, instantiated via Engram, a module that modernizes classic $N$-gram embedding for O(1) lookup. By formulating the Sparsity Allocation problem, we uncover a U-shaped scaling law that optimizes the trade-off between neural computation (MoE) and static memory (Engram). Guided by this law, we scale Engram to 27B par","authors_text":"Bingxuan Wang, Damai Dai, Dongyan Zhao, Han Zhang, Huishuai Zhang, Kezhao Huang, Qinyu Chen, Wangding Zeng, Wenfeng Liang, Xin Cheng, Xingkai Yu, Yukun Li, Zhenda Xie, Zhewen Hao","cross_cats":["cs.AI"],"headline":"Engram introduces conditional memory as a new sparsity axis that lets large language models perform direct O(1) knowledge lookups instead of computing retrieval.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-01-12T09:54:49Z","title":"Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.07372","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-16T04:50:55.810210Z","id":"769f63b4-a10e-4d8a-bec5-7c7cab3fbd4e","model_set":{"reader":"grok-4.3"},"one_line_summary":"Engram adds conditional memory via scalable lookup to LLMs, outperforming iso-parameter MoE baselines on reasoning and long-context tasks by following a U-shaped scaling law for allocating between computation and memory.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Engram introduces conditional memory as a new sparsity axis that lets large language models perform direct O(1) knowledge lookups instead of computing retrieval.","strongest_claim":"Scaling Engram to 27B parameters achieves superior performance over a strictly iso-parameter and iso-FLOPs MoE baseline, with notable gains in reasoning (BBH +5.0, ARC-Challenge +3.7) and long-context retrieval (Multi-Query NIAH: 84.2 to 97.0).","weakest_assumption":"The U-shaped scaling law for sparsity allocation between MoE computation and Engram memory generalizes beyond the tested model sizes and tasks, and the observed mechanistic benefits (relieving early layers, freeing attention) are causally due to the memory module rather than confounding factors in the experimental setup."}},"verdict_id":"769f63b4-a10e-4d8a-bec5-7c7cab3fbd4e"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:be75dca35f9d99eee16a31bfa6df33f463487b20a5a885225029b4203978d577","target":"record","created_at":"2026-05-17T23:38:49Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"d190569143980c41c63739ea5ab93edafb302492a5bf8dd424afdb714a6040b2","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-01-12T09:54:49Z","title_canon_sha256":"f356246ba4b44007608f772ef51593fefac15a455acd4eda5c5b20e0d67de9a2"},"schema_version":"1.0","source":{"id":"2601.07372","kind":"arxiv","version":1}},"canonical_sha256":"839f9b683ac170cd0fd870a4d90bb8a3a891d73cca8c323b9994b23576dabaee","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"839f9b683ac170cd0fd870a4d90bb8a3a891d73cca8c323b9994b23576dabaee","first_computed_at":"2026-05-17T23:38:49.048549Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:38:49.048549Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"t4RVKKFyaXX+eMXfD20S3/DBu7wi2LxxZuTVN8tadK5jrYX+wUALzG9oOnqL8YMiRAhHFhPrP1RYVZyiDA5EAA==","signature_status":"signed_v1","signed_at":"2026-05-17T23:38:49.049202Z","signed_message":"canonical_sha256_bytes"},"source_id":"2601.07372","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:be75dca35f9d99eee16a31bfa6df33f463487b20a5a885225029b4203978d577","sha256:2cfef55bceced902d01380035e847c16e8f968dfbc2445f1000386902bfc9e63"],"state_sha256":"ee6edc5a2ddaee86abd1c19474e8f6394a04a1b6f72d80a23f587c4f3a20d227"}