{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:WAU53BJ3FHRUEMRXOXC2Z6CBJP","short_pith_number":"pith:WAU53BJ3","schema_version":"1.0","canonical_sha256":"b029dd853b29e342323775c5acf8414be96eaaa64a87b2075006dfa4f8ded2f5","source":{"kind":"arxiv","id":"2606.08347","version":1},"attestation_state":"computed","paper":{"title":"Tensorizing Engram: Sharing Latents Across N-Gram Embeddings is Beneficial in LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Danilo Mandic, Giorgos Iacovides, Qibin Zhao, Wuyang Zhou, Yuning Qiu, Yuxuan Gu","submitted_at":"2026-06-06T21:36:51Z","abstract_excerpt":"Modern language models represent text using discrete token-level embeddings, which forces recurring multi-token patterns to be learned implicitly across Transformer layers. Both Over-tokenized Transformers and Engram attempt to address this limitation by explicitly incorporating multi-token (n-gram) memories. However, they rely on separate hash tables for each n-gram order, which introduces hash collisions and prevents nested n-grams from sharing the underlying latent structures. To address these issues, we propose Tensorized Engram (TN-gram), a compact memory module that represents tensorized"},"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":"2606.08347","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-06T21:36:51Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"10be8e90ac67c3de14b98650820cf9224cc57a66ccc654546007453ecfc18126","abstract_canon_sha256":"4e1e55ae33c54f5915e0cd12400d50f38b878bca89f5bab1742169ec5ac85a0f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T01:05:34.297126Z","signature_b64":"9ySf29I6fegxldEEK0e+0hBzx5G7KPu2mCV1lLmansd3TLOuW4b7FvUv6E43FADElJC7EYkrFqjvaXCVGlmlAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b029dd853b29e342323775c5acf8414be96eaaa64a87b2075006dfa4f8ded2f5","last_reissued_at":"2026-06-09T01:05:34.296475Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T01:05:34.296475Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Tensorizing Engram: Sharing Latents Across N-Gram Embeddings is Beneficial in LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Danilo Mandic, Giorgos Iacovides, Qibin Zhao, Wuyang Zhou, Yuning Qiu, Yuxuan Gu","submitted_at":"2026-06-06T21:36:51Z","abstract_excerpt":"Modern language models represent text using discrete token-level embeddings, which forces recurring multi-token patterns to be learned implicitly across Transformer layers. Both Over-tokenized Transformers and Engram attempt to address this limitation by explicitly incorporating multi-token (n-gram) memories. However, they rely on separate hash tables for each n-gram order, which introduces hash collisions and prevents nested n-grams from sharing the underlying latent structures. To address these issues, we propose Tensorized Engram (TN-gram), a compact memory module that represents tensorized"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.08347","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/2606.08347/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":"2606.08347","created_at":"2026-06-09T01:05:34.296607+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.08347v1","created_at":"2026-06-09T01:05:34.296607+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.08347","created_at":"2026-06-09T01:05:34.296607+00:00"},{"alias_kind":"pith_short_12","alias_value":"WAU53BJ3FHRU","created_at":"2026-06-09T01:05:34.296607+00:00"},{"alias_kind":"pith_short_16","alias_value":"WAU53BJ3FHRUEMRX","created_at":"2026-06-09T01:05:34.296607+00:00"},{"alias_kind":"pith_short_8","alias_value":"WAU53BJ3","created_at":"2026-06-09T01:05:34.296607+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/WAU53BJ3FHRUEMRXOXC2Z6CBJP","json":"https://pith.science/pith/WAU53BJ3FHRUEMRXOXC2Z6CBJP.json","graph_json":"https://pith.science/api/pith-number/WAU53BJ3FHRUEMRXOXC2Z6CBJP/graph.json","events_json":"https://pith.science/api/pith-number/WAU53BJ3FHRUEMRXOXC2Z6CBJP/events.json","paper":"https://pith.science/paper/WAU53BJ3"},"agent_actions":{"view_html":"https://pith.science/pith/WAU53BJ3FHRUEMRXOXC2Z6CBJP","download_json":"https://pith.science/pith/WAU53BJ3FHRUEMRXOXC2Z6CBJP.json","view_paper":"https://pith.science/paper/WAU53BJ3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.08347&json=true","fetch_graph":"https://pith.science/api/pith-number/WAU53BJ3FHRUEMRXOXC2Z6CBJP/graph.json","fetch_events":"https://pith.science/api/pith-number/WAU53BJ3FHRUEMRXOXC2Z6CBJP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WAU53BJ3FHRUEMRXOXC2Z6CBJP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WAU53BJ3FHRUEMRXOXC2Z6CBJP/action/storage_attestation","attest_author":"https://pith.science/pith/WAU53BJ3FHRUEMRXOXC2Z6CBJP/action/author_attestation","sign_citation":"https://pith.science/pith/WAU53BJ3FHRUEMRXOXC2Z6CBJP/action/citation_signature","submit_replication":"https://pith.science/pith/WAU53BJ3FHRUEMRXOXC2Z6CBJP/action/replication_record"}},"created_at":"2026-06-09T01:05:34.296607+00:00","updated_at":"2026-06-09T01:05:34.296607+00:00"}