{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:Q3E2PYOCPVNWAKRS2GDFZWHN42","short_pith_number":"pith:Q3E2PYOC","schema_version":"1.0","canonical_sha256":"86c9a7e1c27d5b602a32d1865cd8ede6bd9d081e9f9b4b7dfad34390b72a4c67","source":{"kind":"arxiv","id":"2402.00976","version":4},"attestation_state":"computed","paper":{"title":"Investigating Recurrent Transformers with Dynamic Halt","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.NE"],"primary_cat":"cs.LG","authors_text":"Cornelia Caragea, Jishnu Ray Chowdhury","submitted_at":"2024-02-01T19:47:31Z","abstract_excerpt":"In this paper, we comprehensively study the inductive biases of two major approaches to augmenting Transformers with a recurrent mechanism: (1) the approach of incorporating a depth-wise recurrence similar to Universal Transformers; and (2) the approach of incorporating a chunk-wise temporal recurrence like Temporal Latent Bottleneck. Furthermore, we propose and investigate novel ways to extend and combine the above methods - for example, we propose a global mean-based dynamic halting mechanism for Universal Transformers and an augmentation of Temporal Latent Bottleneck with elements from Univ"},"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":"2402.00976","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-01T19:47:31Z","cross_cats_sorted":["cs.AI","cs.NE"],"title_canon_sha256":"9d7cfbb423e1c8fc9a5379b9ff4f8d08a734258e5da2f6fda6987b612bda7245","abstract_canon_sha256":"ba18f0cea4bfe5762ab6e9d6322bf37d7cdc27b0fd424744905d2e5c1ac9d5f3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:02:43.602162Z","signature_b64":"A13CHzS7xTpF9FNU8GNOWwpvFP7MW3OMdiZVHY5+hj/XaC3PdD/Ppr5GMcVf6fmwlIiHMdwNHNoLytSenF0OBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"86c9a7e1c27d5b602a32d1865cd8ede6bd9d081e9f9b4b7dfad34390b72a4c67","last_reissued_at":"2026-07-05T10:02:43.601719Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:02:43.601719Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Investigating Recurrent Transformers with Dynamic Halt","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.NE"],"primary_cat":"cs.LG","authors_text":"Cornelia Caragea, Jishnu Ray Chowdhury","submitted_at":"2024-02-01T19:47:31Z","abstract_excerpt":"In this paper, we comprehensively study the inductive biases of two major approaches to augmenting Transformers with a recurrent mechanism: (1) the approach of incorporating a depth-wise recurrence similar to Universal Transformers; and (2) the approach of incorporating a chunk-wise temporal recurrence like Temporal Latent Bottleneck. Furthermore, we propose and investigate novel ways to extend and combine the above methods - for example, we propose a global mean-based dynamic halting mechanism for Universal Transformers and an augmentation of Temporal Latent Bottleneck with elements from Univ"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.00976","kind":"arxiv","version":4},"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/2402.00976/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":"2402.00976","created_at":"2026-07-05T10:02:43.601775+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.00976v4","created_at":"2026-07-05T10:02:43.601775+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.00976","created_at":"2026-07-05T10:02:43.601775+00:00"},{"alias_kind":"pith_short_12","alias_value":"Q3E2PYOCPVNW","created_at":"2026-07-05T10:02:43.601775+00:00"},{"alias_kind":"pith_short_16","alias_value":"Q3E2PYOCPVNWAKRS","created_at":"2026-07-05T10:02:43.601775+00:00"},{"alias_kind":"pith_short_8","alias_value":"Q3E2PYOC","created_at":"2026-07-05T10:02:43.601775+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.20613","citing_title":"HRM-Text: Efficient Pretraining Beyond Scaling","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Q3E2PYOCPVNWAKRS2GDFZWHN42","json":"https://pith.science/pith/Q3E2PYOCPVNWAKRS2GDFZWHN42.json","graph_json":"https://pith.science/api/pith-number/Q3E2PYOCPVNWAKRS2GDFZWHN42/graph.json","events_json":"https://pith.science/api/pith-number/Q3E2PYOCPVNWAKRS2GDFZWHN42/events.json","paper":"https://pith.science/paper/Q3E2PYOC"},"agent_actions":{"view_html":"https://pith.science/pith/Q3E2PYOCPVNWAKRS2GDFZWHN42","download_json":"https://pith.science/pith/Q3E2PYOCPVNWAKRS2GDFZWHN42.json","view_paper":"https://pith.science/paper/Q3E2PYOC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.00976&json=true","fetch_graph":"https://pith.science/api/pith-number/Q3E2PYOCPVNWAKRS2GDFZWHN42/graph.json","fetch_events":"https://pith.science/api/pith-number/Q3E2PYOCPVNWAKRS2GDFZWHN42/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Q3E2PYOCPVNWAKRS2GDFZWHN42/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Q3E2PYOCPVNWAKRS2GDFZWHN42/action/storage_attestation","attest_author":"https://pith.science/pith/Q3E2PYOCPVNWAKRS2GDFZWHN42/action/author_attestation","sign_citation":"https://pith.science/pith/Q3E2PYOCPVNWAKRS2GDFZWHN42/action/citation_signature","submit_replication":"https://pith.science/pith/Q3E2PYOCPVNWAKRS2GDFZWHN42/action/replication_record"}},"created_at":"2026-07-05T10:02:43.601775+00:00","updated_at":"2026-07-05T10:02:43.601775+00:00"}