{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RKEVB46SLTGGRKKE3M6SPDYFY6","short_pith_number":"pith:RKEVB46S","schema_version":"1.0","canonical_sha256":"8a8950f3d25ccc68a944db3d278f05c794e4ccb3eb34ebf386cec108785b38d2","source":{"kind":"arxiv","id":"2506.17671","version":2},"attestation_state":"computed","paper":{"title":"TPTT: Transforming Pretrained Transformers into Titans","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Fabien Furfaro","submitted_at":"2025-06-21T10:06:07Z","abstract_excerpt":"Transformer-based large language models (LLMs) have achieved strong performance across many natural language processing tasks. Nonetheless, their quadratic computational and memory requirements, particularly in self-attention layers, pose challenges for efficient inference on long contexts and for deployment in resource-limited environments. We present TPTT (Transforming Pretrained Transformers into Titans), a framework designed to augment pretrained Transformers with linearized attention (LiZA) and internal memory gating via Memory as Gate (MaG), applied without full retraining. TPTT supports"},"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":"2506.17671","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-21T10:06:07Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"8bc2318f1e916de8db26785e83237da0d3ce26bc8bde5b2d22451a22a324643b","abstract_canon_sha256":"bbca8f66cfd2b9eda17e012b250258a733cc32aacbd399543b714a9124352207"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:03:27.834802Z","signature_b64":"XtuDVukaW5ILk5VmjTusCldPUtYWpjAV3xmNlAooSN8NxbcLq3hO88Mayl/fpOfhYBBCqHHTPAxEWJnH5PFbDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8a8950f3d25ccc68a944db3d278f05c794e4ccb3eb34ebf386cec108785b38d2","last_reissued_at":"2026-07-05T12:03:27.834299Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:03:27.834299Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TPTT: Transforming Pretrained Transformers into Titans","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Fabien Furfaro","submitted_at":"2025-06-21T10:06:07Z","abstract_excerpt":"Transformer-based large language models (LLMs) have achieved strong performance across many natural language processing tasks. Nonetheless, their quadratic computational and memory requirements, particularly in self-attention layers, pose challenges for efficient inference on long contexts and for deployment in resource-limited environments. We present TPTT (Transforming Pretrained Transformers into Titans), a framework designed to augment pretrained Transformers with linearized attention (LiZA) and internal memory gating via Memory as Gate (MaG), applied without full retraining. TPTT supports"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.17671","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/2506.17671/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":"2506.17671","created_at":"2026-07-05T12:03:27.834364+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.17671v2","created_at":"2026-07-05T12:03:27.834364+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.17671","created_at":"2026-07-05T12:03:27.834364+00:00"},{"alias_kind":"pith_short_12","alias_value":"RKEVB46SLTGG","created_at":"2026-07-05T12:03:27.834364+00:00"},{"alias_kind":"pith_short_16","alias_value":"RKEVB46SLTGGRKKE","created_at":"2026-07-05T12:03:27.834364+00:00"},{"alias_kind":"pith_short_8","alias_value":"RKEVB46S","created_at":"2026-07-05T12:03:27.834364+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.16350","citing_title":"Federated Nested Learning: Collaborative Training of Self-Referential Memories for Test-Time Adaptation","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RKEVB46SLTGGRKKE3M6SPDYFY6","json":"https://pith.science/pith/RKEVB46SLTGGRKKE3M6SPDYFY6.json","graph_json":"https://pith.science/api/pith-number/RKEVB46SLTGGRKKE3M6SPDYFY6/graph.json","events_json":"https://pith.science/api/pith-number/RKEVB46SLTGGRKKE3M6SPDYFY6/events.json","paper":"https://pith.science/paper/RKEVB46S"},"agent_actions":{"view_html":"https://pith.science/pith/RKEVB46SLTGGRKKE3M6SPDYFY6","download_json":"https://pith.science/pith/RKEVB46SLTGGRKKE3M6SPDYFY6.json","view_paper":"https://pith.science/paper/RKEVB46S","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.17671&json=true","fetch_graph":"https://pith.science/api/pith-number/RKEVB46SLTGGRKKE3M6SPDYFY6/graph.json","fetch_events":"https://pith.science/api/pith-number/RKEVB46SLTGGRKKE3M6SPDYFY6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RKEVB46SLTGGRKKE3M6SPDYFY6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RKEVB46SLTGGRKKE3M6SPDYFY6/action/storage_attestation","attest_author":"https://pith.science/pith/RKEVB46SLTGGRKKE3M6SPDYFY6/action/author_attestation","sign_citation":"https://pith.science/pith/RKEVB46SLTGGRKKE3M6SPDYFY6/action/citation_signature","submit_replication":"https://pith.science/pith/RKEVB46SLTGGRKKE3M6SPDYFY6/action/replication_record"}},"created_at":"2026-07-05T12:03:27.834364+00:00","updated_at":"2026-07-05T12:03:27.834364+00:00"}