mlsanityData to 7 Oct 2026

Core SSM architectures for language

278 papers0 this weekHigh confidence in this grouping: the papers form one clear topicHigh confidence in this grouping: the papers form one clear topic · About

Keywords: mamba, state, ssms, language, sequence

Key facts. Core SSM architectures for language: 278 papers on arXiv since 2021 (data through 7 Oct 2026). 152 of its 278 mapped papers (55%) cite “Mamba: Linear-Time Sequence Modeling with Selective State Spaces” (2023, arXiv 2312.00752). 114 of its 278 mapped papers (41%) cite “Efficiently Modeling Long Sequences with Structured State Spaces” (2021, arXiv 2111.00396). 8 new papers 18 Sep–7 Oct.

What’s new

18 Sep – 7 Oct 2026

8 new papers, 18 Sep – 7 Oct; 0 this week.

  1. Earlier · 8
  2. Distributed Learning with Selective State Space Models: Architecture-Aware Convergence Analysis
  3. Low-Discrepancy Dither for Quantized Recurrent State Caches
  4. SketchSSM: Write to the Full State, Read from a Compact Sketch
  5. Elastic Selective Spectral Hybrids for Train-Once, Export-Many Budgeted Inference
  6. Residual Streams Read, Recurrent States Remember: The Global Workspace in Mamba Models
  7. Non-Commutative State Tracking with Input-Dependent Low-Rank Updates in Mamba-3
  8. Comparing Latent Concept Formation in State Space Models and Transformers via Sparse Autoencoders
  9. H-Spec: Parallel Speculative Decoding Without a Drafter-Side KV Cache

Foundations

Most cited by papers here

  1. Mamba: Linear-Time Sequence Modeling with Selective State Spaces2023cited by 152 here
  2. Efficiently Modeling Long Sequences with Structured State Spaces2021cited by 114 here
  3. Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Duality2024cited by 99 here
  4. Simplified State Space Layers for Sequence Modeling2022cited by 53 here
  5. Jamba: A Hybrid Transformer-Mamba Language Model2024cited by 40 here

Lineage

The papers this direction builds on, oldest first.

Each paper on the line cites the one before.

  1. 2021Efficiently Modeling Long Sequences with Structured State Spacescited by 114 here
  2. 2022Diagonal State Spaces are as Effective as Structured State Spacescited by 33 here
  3. 2023Mamba: Linear-Time Sequence Modeling with Selective State Spacescited by 152 here
  4. 2024U-Mamba: Enhancing Long-range Dependency for Biomedical Image Segmentationcited by 11 here
  5. 2024Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Modelcited by 33 here
  6. 2024Graph-Mamba: Towards Long-Range Graph Sequence Modeling with Selective State Spacescited by 13 here
  7. 2024The Hidden Attention of Mamba Modelscited by 25 here
  8. 2024Jamba: A Hybrid Transformer-Mamba Language Modelcited by 40 here
  9. 2024Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Dualitycited by 99 here
  10. 2024An Empirical Study of Mamba-based Language Modelscited by 30 here
  11. 2024Falcon Mamba: The First Competitive Attention-free 7B Language Modelcited by 12 here
13 more it builds on, not in this chain
  1. 2022How to Train Your HiPPO: State Space Models with Generalized Orthogonal Basis Projectionscited by 12 here
  2. 2022Simplified State Space Layers for Sequence Modelingcited by 53 here
  3. 2023Gated Linear Attention Transformers with Hardware-Efficient Trainingcited by 23 here
  4. 2024MoE-Mamba: Efficient Selective State Space Models with Mixture of Expertscited by 14 here
  5. 2024MambaByte: Token-free Selective State Space Modelcited by 12 here
  6. 2024BlackMamba: Mixture of Experts for State-Space Modelscited by 11 here
  7. 2024Can Mamba Learn How to Learn? A Comparative Study on In-Context Learning Taskscited by 22 here
  8. 2024Mamba-ND: Selective State Space Modeling for Multi-Dimensional Datacited by 10 here
  9. 2024Theoretical Foundations of Deep Selective State-Space Modelscited by 13 here
  10. 2024VideoMamba: State Space Model for Efficient Video Understandingcited by 13 here
  11. 2024Samba: Simple Hybrid State Space Models for Efficient Unlimited Context Language Modelingcited by 16 here
  12. 2024DeciMamba: Exploring the Length Extrapolation Potential of Mambacited by 14 here
  13. 2024The Mamba in the Llama: Distilling and Accelerating Hybrid Modelscited by 11 here

The 24 most cited of 258.

Browse

30 of 278, ranked 6–35

  1. 6Diagonal State Spaces are as Effective as Structured State Spaces202233
  2. 7An Empirical Study of Mamba-based Language Models202430
  3. 8The Hidden Attention of Mamba Models202425
  4. 9Gated Linear Attention Transformers with Hardware-Efficient Training202323
  5. 10Can Mamba Learn How to Learn? A Comparative Study on In-Context Learning Tasks202422
  6. 11Samba: Simple Hybrid State Space Models for Efficient Unlimited Context Language Modeling202416
  7. 12MoE-Mamba: Efficient Selective State Space Models with Mixture of Experts202414
  8. 13DeciMamba: Exploring the Length Extrapolation Potential of Mamba202414
  9. 14Theoretical Foundations of Deep Selective State-Space Models202413
  10. 15MambaByte: Token-free Selective State Space Model202412
All 30
  1. 16Falcon Mamba: The First Competitive Attention-free 7B Language Model202412
  2. 17How to Train Your HiPPO: State Space Models with Generalized Orthogonal Basis Projections202212
  3. 18BlackMamba: Mixture of Experts for State-Space Models202411
  4. 19The Mamba in the Llama: Distilling and Accelerating Hybrid Models202411
  5. 20Mamba-ND: Selective State Space Modeling for Multi-Dimensional Data202410
  6. 21Simple linear attention language models balance the recall-throughput tradeoff202410
  7. 22Mamba-360: Survey of State Space Models as Transformer Alternative for Long Sequence Modelling: Methods, Applications, and Challenges202410
  8. 23Is Mamba Capable of In-Context Learning?20249
  9. 24Transformers to SSMs: Distilling Quadratic Knowledge to Subquadratic Models20249
  10. 25Structured State Space Models for In-Context Reinforcement Learning20239
  11. 26Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Models20257
  12. 27LongMamba: Enhancing Mamba's Long Context Capabilities via Training-Free Receptive Field Enlargement20257
  13. 28MambaLRP: Explaining Selective State Space Sequence Models20246
  14. 29Jamba-1.5: Hybrid Transformer-Mamba Models at Scale20246
  15. 30From Markov to Laplace: How Mamba In-Context Learns Markov Chains20255
  16. 31Llamba: Scaling Distilled Recurrent Models for Efficient Language Processing20255
  17. 32MambaTab: A Plug-and-Play Model for Learning Tabular Data20245
  18. 33Locating and Editing Factual Associations in Mamba20245
  19. 34Unlocking State-Tracking in Linear RNNs Through Negative Eigenvalues20245
  20. 35Understanding and Mitigating Bottlenecks of State Space Models through the Lens of Recency and Over-smoothing20245