{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:T2U3PZRN663Y2MHSCP32MFYS67","short_pith_number":"pith:T2U3PZRN","schema_version":"1.0","canonical_sha256":"9ea9b7e62df7b78d30f213f7a61712f7e4afd01ec1212da472fefca311b0c786","source":{"kind":"arxiv","id":"2406.07522","version":3},"attestation_state":"computed","paper":{"title":"Samba: Simple Hybrid State Space Models for Efficient Unlimited Context Language Modeling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Chen Liang, Liliang Ren, Weizhu Chen, Yadong Lu, Yang Liu, Yelong Shen","submitted_at":"2024-06-11T17:50:51Z","abstract_excerpt":"Efficiently modeling sequences with infinite context length has long been a challenging problem. Previous approaches have either suffered from quadratic computational complexity or limited extrapolation ability in length generalization. In this work, we present Samba, a simple hybrid architecture that layer-wise combines Mamba, a selective State Space Model (SSM), with Sliding Window Attention (SWA). Samba selectively compresses a given sequence into recurrent hidden states while still maintaining the ability to precisely recall recent memories with the attention mechanism. We scale Samba up t"},"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":"2406.07522","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-11T17:50:51Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"f2aab5e753390dcf0e92270165e918167defc85ae4e7f8385912a3d08f710de2","abstract_canon_sha256":"7c7299865e3890dbe8e9b01835c3cfdea686287a8fd4183e6f4b0a144d8d3a49"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:21:20.484516Z","signature_b64":"epigN7ZFtkJkBg8GmUaDYZbMSrsEKKnZNkt7cwOl/+GqxIvSMFAJu2u5qXek0QGPZgEGi4FjVAt5+918ROxpCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9ea9b7e62df7b78d30f213f7a61712f7e4afd01ec1212da472fefca311b0c786","last_reissued_at":"2026-07-05T10:21:20.483982Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:21:20.483982Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Samba: Simple Hybrid State Space Models for Efficient Unlimited Context Language Modeling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Chen Liang, Liliang Ren, Weizhu Chen, Yadong Lu, Yang Liu, Yelong Shen","submitted_at":"2024-06-11T17:50:51Z","abstract_excerpt":"Efficiently modeling sequences with infinite context length has long been a challenging problem. Previous approaches have either suffered from quadratic computational complexity or limited extrapolation ability in length generalization. In this work, we present Samba, a simple hybrid architecture that layer-wise combines Mamba, a selective State Space Model (SSM), with Sliding Window Attention (SWA). Samba selectively compresses a given sequence into recurrent hidden states while still maintaining the ability to precisely recall recent memories with the attention mechanism. We scale Samba up t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.07522","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/2406.07522/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":"2406.07522","created_at":"2026-07-05T10:21:20.484054+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.07522v3","created_at":"2026-07-05T10:21:20.484054+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.07522","created_at":"2026-07-05T10:21:20.484054+00:00"},{"alias_kind":"pith_short_12","alias_value":"T2U3PZRN663Y","created_at":"2026-07-05T10:21:20.484054+00:00"},{"alias_kind":"pith_short_16","alias_value":"T2U3PZRN663Y2MHS","created_at":"2026-07-05T10:21:20.484054+00:00"},{"alias_kind":"pith_short_8","alias_value":"T2U3PZRN","created_at":"2026-07-05T10:21:20.484054+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":21,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08032","citing_title":"What to Keep, What to Forget: A Rate--Distortion View of Memory Compaction in LLMs and Agents","ref_index":101,"is_internal_anchor":true},{"citing_arxiv_id":"2606.12895","citing_title":"LongSpike: Fractional Order Spiking State Space Models for Efficient Long Sequence Learning","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02332","citing_title":"Forget Attention: Importance-Aware Attention Is All You Need","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26099","citing_title":"Do Language Models Need Sleep? Offline Recurrence for Improved Online Inference","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07604","citing_title":"Contribution Weights: A Geometrical Analysis of Self-Attention Transformers","ref_index":140,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03014","citing_title":"MOSAIC: Efficient Mixture-of-Agent Scheduling via Adaptive Aggregation and Inference Concurrency","ref_index":73,"is_internal_anchor":false},{"citing_arxiv_id":"2408.01129","citing_title":"A Survey of Mamba","ref_index":156,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22416","citing_title":"Asymmetric Virtual Memory Paging for Hybrid Mamba-Transformer Inference","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12640","citing_title":"MambaPanoptic: A Vision Mamba-based Structured State Space Framework for Panoptic Segmentation","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2506.13183","citing_title":"MT-PCR: Hybrid Mamba-Transformer Network with Spatial Serialization for Point Cloud Registration","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2509.24552","citing_title":"Short window attention enables long-term memorization","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2510.04800","citing_title":"Hybrid Architectures for Language Models: Systematic Analysis and Design Insights","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2510.26083","citing_title":"Nirvana: A Specialized Generalist Model With Task-Aware Memory Mechanism","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2601.01972","citing_title":"Hidden State Poisoning Attacks against Mamba-based Language Models","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2603.14360","citing_title":"M$^2$RNN: Non-Linear RNNs with Matrix-Valued States for Scalable Language Modeling","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2501.00663","citing_title":"Titans: Learning to Memorize at Test Time","ref_index":89,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12640","citing_title":"MambaPanoptic: A Vision Mamba-based Structured State Space Framework for Panoptic Segmentation","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2506.13585","citing_title":"MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24715","citing_title":"Long-Context Aware Upcycling: A New Frontier for Hybrid LLM Scaling","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05066","citing_title":"The Impossibility Triangle of Long-Context Modeling","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01106","citing_title":"Component-Aware Self-Speculative Decoding in Hybrid Language Models","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/T2U3PZRN663Y2MHSCP32MFYS67","json":"https://pith.science/pith/T2U3PZRN663Y2MHSCP32MFYS67.json","graph_json":"https://pith.science/api/pith-number/T2U3PZRN663Y2MHSCP32MFYS67/graph.json","events_json":"https://pith.science/api/pith-number/T2U3PZRN663Y2MHSCP32MFYS67/events.json","paper":"https://pith.science/paper/T2U3PZRN"},"agent_actions":{"view_html":"https://pith.science/pith/T2U3PZRN663Y2MHSCP32MFYS67","download_json":"https://pith.science/pith/T2U3PZRN663Y2MHSCP32MFYS67.json","view_paper":"https://pith.science/paper/T2U3PZRN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.07522&json=true","fetch_graph":"https://pith.science/api/pith-number/T2U3PZRN663Y2MHSCP32MFYS67/graph.json","fetch_events":"https://pith.science/api/pith-number/T2U3PZRN663Y2MHSCP32MFYS67/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/T2U3PZRN663Y2MHSCP32MFYS67/action/timestamp_anchor","attest_storage":"https://pith.science/pith/T2U3PZRN663Y2MHSCP32MFYS67/action/storage_attestation","attest_author":"https://pith.science/pith/T2U3PZRN663Y2MHSCP32MFYS67/action/author_attestation","sign_citation":"https://pith.science/pith/T2U3PZRN663Y2MHSCP32MFYS67/action/citation_signature","submit_replication":"https://pith.science/pith/T2U3PZRN663Y2MHSCP32MFYS67/action/replication_record"}},"created_at":"2026-07-05T10:21:20.484054+00:00","updated_at":"2026-07-05T10:21:20.484054+00:00"}