{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:LLFKJLEZXAGG6RQ2TP4NOGJWZK","short_pith_number":"pith:LLFKJLEZ","schema_version":"1.0","canonical_sha256":"5acaa4ac99b80c6f461a9bf8d71936caac61d37e97fe3aee8ca4643079dd598f","source":{"kind":"arxiv","id":"2406.14528","version":3},"attestation_state":"computed","paper":{"title":"DeciMamba: Exploring the Length Extrapolation Potential of Mamba","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Amir Globerson, Assaf Ben-Kish, Itamar Zimerman, Lior Wolf, Nadav Cohen, Raja Giryes, Shady Abu-Hussein","submitted_at":"2024-06-20T17:40:18Z","abstract_excerpt":"Long-range sequence processing poses a significant challenge for Transformers due to their quadratic complexity in input length. A promising alternative is Mamba, which demonstrates high performance and achieves Transformer-level capabilities while requiring substantially fewer computational resources. In this paper we explore the length-generalization capabilities of Mamba, which we find to be relatively limited. Through a series of visualizations and analyses we identify that the limitations arise from a restricted effective receptive field, dictated by the sequence length used during traini"},"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":"2406.14528","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-20T17:40:18Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6f94bc2ab4b062dacbb8a6524975a53a6644ca755c0b8da7b050decc6a5b5d68","abstract_canon_sha256":"99b9d71814ce6b0ed0b2fd845f83c7590d5f2fcb842938580a86f9f16a2e66c9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:46:55.611738Z","signature_b64":"/oyUTlzX5vsVW7CU1Av0PHOVoSGtsvHll59bpMchsl7lsuMnlQlnegJqiVHWRme1TPyz3PObGUSFvQZ1YVtqBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5acaa4ac99b80c6f461a9bf8d71936caac61d37e97fe3aee8ca4643079dd598f","last_reissued_at":"2026-07-05T10:46:55.611240Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:46:55.611240Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DeciMamba: Exploring the Length Extrapolation Potential of Mamba","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Amir Globerson, Assaf Ben-Kish, Itamar Zimerman, Lior Wolf, Nadav Cohen, Raja Giryes, Shady Abu-Hussein","submitted_at":"2024-06-20T17:40:18Z","abstract_excerpt":"Long-range sequence processing poses a significant challenge for Transformers due to their quadratic complexity in input length. A promising alternative is Mamba, which demonstrates high performance and achieves Transformer-level capabilities while requiring substantially fewer computational resources. In this paper we explore the length-generalization capabilities of Mamba, which we find to be relatively limited. Through a series of visualizations and analyses we identify that the limitations arise from a restricted effective receptive field, dictated by the sequence length used during traini"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.14528","kind":"arxiv","version":3},"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/2406.14528/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":"2406.14528","created_at":"2026-07-05T10:46:55.611301+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.14528v3","created_at":"2026-07-05T10:46:55.611301+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.14528","created_at":"2026-07-05T10:46:55.611301+00:00"},{"alias_kind":"pith_short_12","alias_value":"LLFKJLEZXAGG","created_at":"2026-07-05T10:46:55.611301+00:00"},{"alias_kind":"pith_short_16","alias_value":"LLFKJLEZXAGG6RQ2","created_at":"2026-07-05T10:46:55.611301+00:00"},{"alias_kind":"pith_short_8","alias_value":"LLFKJLEZ","created_at":"2026-07-05T10:46:55.611301+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2510.04800","citing_title":"Hybrid Architectures for Language Models: Systematic Analysis and Design Insights","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2509.26645","citing_title":"TTT3R: 3D Reconstruction as Test-Time Training","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.16424","citing_title":"Safety, Security, and Cognitive Risks in State-Space Models: A Systematic Threat Analysis with Spectral, Stateful, and Capacity Attacks","ref_index":45,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LLFKJLEZXAGG6RQ2TP4NOGJWZK","json":"https://pith.science/pith/LLFKJLEZXAGG6RQ2TP4NOGJWZK.json","graph_json":"https://pith.science/api/pith-number/LLFKJLEZXAGG6RQ2TP4NOGJWZK/graph.json","events_json":"https://pith.science/api/pith-number/LLFKJLEZXAGG6RQ2TP4NOGJWZK/events.json","paper":"https://pith.science/paper/LLFKJLEZ"},"agent_actions":{"view_html":"https://pith.science/pith/LLFKJLEZXAGG6RQ2TP4NOGJWZK","download_json":"https://pith.science/pith/LLFKJLEZXAGG6RQ2TP4NOGJWZK.json","view_paper":"https://pith.science/paper/LLFKJLEZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.14528&json=true","fetch_graph":"https://pith.science/api/pith-number/LLFKJLEZXAGG6RQ2TP4NOGJWZK/graph.json","fetch_events":"https://pith.science/api/pith-number/LLFKJLEZXAGG6RQ2TP4NOGJWZK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LLFKJLEZXAGG6RQ2TP4NOGJWZK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LLFKJLEZXAGG6RQ2TP4NOGJWZK/action/storage_attestation","attest_author":"https://pith.science/pith/LLFKJLEZXAGG6RQ2TP4NOGJWZK/action/author_attestation","sign_citation":"https://pith.science/pith/LLFKJLEZXAGG6RQ2TP4NOGJWZK/action/citation_signature","submit_replication":"https://pith.science/pith/LLFKJLEZXAGG6RQ2TP4NOGJWZK/action/replication_record"}},"created_at":"2026-07-05T10:46:55.611301+00:00","updated_at":"2026-07-05T10:46:55.611301+00:00"}