الباحثون

Hongyi Wu

المنشورات 2

نسخة أولية وصول مفتوح

When Trees Are Not Enough: Learning Mixed-Topology Feature Graphs with Adaptive Graph Sparse Autoencoders

Xiaozuo Shen, Yifei Cai, Tian Tan وآخرون · 2026

Sparse autoencoders (SAEs) expose interpretable features in large language model activations, yet existing structured SAEs impose single-parent trees or forests, while post-hoc graphs permit multiple parents but neither guide feature learning nor ensure reliable relation recovery. We introduce the Adaptive Graph Sparse …

نسخة أولية وصول مفتوح

DegreeSpar: Structured Degree Sparsity for Efficient Secure Transformer Inference

Yifei Cai, Zhuoran Li, Xiaozuo Shen وآخرون · 2026

Secure Transformer inference protects sensitive inputs but incurs substantial cryptographic overhead, with nonlinear operations such as Softmax and GeLU becoming major bottlenecks. Existing compression methods reduce nonlinear complexity, sequence-dependent computation, or model structure through separately defined com …

المؤلفون المشاركون