AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

SparseSSM: Efficient Selective Structured State Space Models Can Be Pruned in One-Shot

Kaiwen Tuo, Huan Wang

ICML 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Layer-wise pruning of LLMs · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use SparseSSM time-weighted OBS scoring instead of SparseGPT for A_log pruning at 50% sparsityMamba-370M · WikiText-2perplexity: 2696 -> 15.04Improved
Use SparseSSM instead of Wanda for A_log pruning at 50% sparsityMamba-130M · WikiText-2perplexity: 77.84 -> 21.26Improved
Use SparseSSM instead of SparseGPTMamba-370M · zero-shot average of 5 tasksaverage accuracy (%): 44.56 -> 47.47Improved
Prune A_log of Mamba-2.8B to 50% sparsity with SparseSSMMamba-2.8B · WikiText-2perplexity: 9.45 -> 10.63Mixed
Replace power-law time-weighted aggregation with simple L2 aggregation over time steps (ablation)SparseSSM · Mamba-370M zero-shot accuracyaccuracyImproved
Add sensitivity-aware sparsity allocation (alpha=0.04) to FFN pruning (ablation)SparseSSM on FFN modules · Mamba-370M zero-shot accuracyaccuracyImproved
Use fewer than 16 calibration samples (ablation)SparseSSM · Mamba-370M SSM pruningaccuracyHurt
Use 64 calibration samples (ablation)SparseSSM · Mamba-370M SSM pruningaccuracy and pruning timeImproved
Use N:M (2:4) sparsity with SparseSSM instead of magnitude pruningMamba-370M SSM · WikiText-2perplexity: 77.20 -> 16.90Improved
Structured column pruning of A_log at 50% sparsityMamba-370M SSM · SSM module inferencespeedup: 1.72xImproved
Prune in_proj and out_proj as heavily as other FFN modules (ablation)Mamba FFN pruning · Mamba architecturedegradationHurt
Prune SSM and FFN jointly with SparseSSMfull Mamba · multiple model scalesperplexity and zero-shot accuracyImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy