methods · token mixer
linear attention & state space
18 methods filed at this node or below it, from the sources of 10 models.
In practice
3 features checked against each model’s own code and config.json, apart from its documents. A column’s page has the verdict and the lines.
adopteddiscussednot usedblank: documents silent✓in its code✕not in its code
| Model | Linear-attention or state-space layers alongside full attention | Gated DeltaNet layers | Mamba-2 layers |
|---|---|---|---|
| GLM-5.3-Flash | in its code | in its code | not in its code |
| Qwen3.8-Flash-Next | in its code | in its code | not in its code |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B | in its code | not in its code | in its code |
| Kimi K3 | in its code | in its code | not in its code |
| NVIDIA-Nemotron-3-Ultra-550B-A55B | in its code | not in its code | in its code |
| Qwen3.6-35B-A3B | in its code | in its code | not in its code |
| Qwen3.5-397B-A17B | in its code | in its code | not in its code |
| NVIDIA-Nemotron-3.5-Super-VL-120B-A12B | code unclear | code unclear | in its code |
| Mistral-Large-4 | not in its code | not in its code | not in its code |
| LFM3-1.7B | not in its code | not in its code | not in its code |
| MiniMax-M3.1-Flash | not in its code | not in its code | not in its code |
| DeepSeek-V4.1-Flash-1017 | not in its code | not in its code | not in its code |
| Step-5-Preview | not in its code | not in its code | not in its code |
| Kolibri-1 | not in its code | not in its code | not in its code |
| MiMo-V2.6-Flash | not in its code | not in its code | not in its code |
| MiMo-V2.6-Pro | not in its code | not in its code | not in its code |
| DeepSeek-V4.1-Flash | not in its code | not in its code | not in its code |
| Hy4-preview | not in its code | not in its code | not in its code |
| DeepSeek-V4-Flash-Vision-Exp | not in its code | not in its code | not in its code |
| GLM-5.3 | not in its code | not in its code | not in its code |
| DeepSeek-V4-Pro-0813 | not in its code | not in its code | not in its code |
| DeepSeek-V4-Flash-0731 | not in its code | not in its code | not in its code |
| Xiaomi-Robotics-1 | not in its code | not in its code | not in its code |
| Laguna-S-2.1 | not in its code | not in its code | not in its code |
| Inkling | not in its code | not in its code | not in its code |
| Hy3 | not in its code | not in its code | not in its code |
| GLM-5.2 | not in its code | not in its code | not in its code |
| MiniMax-M3 | not in its code | not in its code | not in its code |
| Step-3.7-Flash | not in its code | not in its code | not in its code |
| MolmoAct 2 | not in its code | not in its code | not in its code |
| DeepSeek-V4-Flash | not in its code | not in its code | not in its code |
| DeepSeek-V4-Pro | not in its code | not in its code | not in its code |
| MiMo-V2.5 | not in its code | not in its code | not in its code |
| MiMo-V2.5-Pro | not in its code | not in its code | not in its code |
| Gemma 4 31B | not in its code | not in its code | not in its code |
| DeepSeek-V3.2 | not in its code | not in its code | not in its code |
| X-VLA | not in its code | not in its code | not in its code |
| gpt-oss-120b | not in its code | not in its code | not in its code |
| OpenVLA / OpenVLA-OFT | not in its code | not in its code | not in its code |
In this branch 18
Everything filed at this node or below it, with one collapsible heading per child node.
filed here 2
1 only mentioned as not used
Mamba 4
By model
Which of this branch's methods each model's own documents describe, and how strongly. Under each model: its strongest treatment anywhere in the branch.