#model pruning

try —

6 papers match

cs.LG2026

Beyond Geometric Complementarity: Coherent Overlap in Sparse Mixture-of-Experts Routing

Huiyuan Tian, Bonan Xu, Shijian Li

The paper investigates how sparse mixture-of-experts language models route tokens to multiple experts, showing that expert subspaces overlap substantially yet routing still selects…

#mixture of experts#sparse routing#language models#expert overlap
eess.AS2026

VAD to the Bone: Ultra-Tiny Speech Activity Detection for Edge Deployment

Stephen Bauer, Sheila Seidel, Shanza Iftikhar +2

The paper introduces kiloVAD, an ultra‑tiny, CNN‑only voice activity detection model designed for edge devices, using standard Mel features, structured pruning with self‑distillati…

#voice activity detection#edge computing#model pruning#quantization-aware training
cs.CV2026

Post-Training Pruning for Diffusion Transformers

Chengzhi Hu, Xuewen Liu, Jing Zhang +3

The paper introduces DiT-Pruning, a post‑training pruning method tailored for Diffusion Transformers that uses a new energy‑based saliency metric and clustering‑aware granularity t…

#diffusion models#transformers#model pruning#post‑training compression
eess.AS2026

Efficient Text-to-Audio Generation via Pruning

Arshdeep Singh, Yi Yuan, Yun Chen +2

The paper applies filter‑based pruning to the U‑Net backbone of the AudioLDM text‑to‑audio diffusion model, reducing most of its parameters and compute while preserving generation…

#text-to-audio generation#model pruning#diffusion models#audio synthesis
cs.CL2026

UMoE:Unlocking Every Expert in Domain-Specific Training

Xuefeng Li, Pengfei Liu

The paper introduces UMoE, a method that prunes low‑saliency experts and regrows new ones to better align a mixture‑of‑experts language model with a target domain before fine‑tunin…

#mixture-of-experts#domain adaptation#model pruning#expert regrowth
cs.LG2026

An Exact Instrument for State Usage in Selective State-Space Models, and the Input-Driven Migration It Reveals

Raktim Bhattacharya

The paper introduces an exact measurement tool that quantifies how selective state‑space models like Mamba use their internal modes, enabling precise prediction of pruning errors a…

#state-space models#model pruning#mode usage analysis#input-dependent dynamics

One search, two signals: results blend meaning (embedding similarity, so papers that never use your words still surface) with keyword matches on titles, abstracts and summaries. Free, no sign-in needed.