S2-MoE: Enabling Efficient Self-Speculative Decoding for Mixture-of-Experts on Edge Devices
Haochen HuangShengxuan QiuMeng Li
Aug 2026
Artificial Intelligence
Abstract
Deploying large language models (LLMs) for inference on edge devices is challenging due to severe memory and bandwidth constraints. While speculative decoding and Mixture-of-Experts (MoE) have been proposed to improve inference efficiency, naively combining them often incurs excessive verification overhead and poor expert reuse, limiting their effectiveness in memory-bound edge settings. In this work, we propose S2-MoE, an efficient self-speculative decoding framework for MoE inference on edge devices. S2-MoE reduces redundant verification through routing-aware adaptive speculative expansion, improves verification efficiency with reuse-aware expert gating, and aligns draft and target execution via shared context. Implemented in llama$.$cpp, S2-MoE achieves up to $5.3\times$ speedup (about $2.0\times$ on average) over standard autoregressive decoding across diverse MoE models and datasets on edge devices. Code is available at https://github.com/angerybob/S2-MoE.
Investigating how experienced developers use agents in building software, including their motivations, strategies, task suitability, and sentiments finds that while experienced developers value agents as a productivity boost, they retain their agency in software design and implementation out of insistence on fundamental software quality attributes.
This systematic review evaluates 55 studies from 2017 to 2023 on the application of machine learning techniques to ASD, highlighting key challenges and opportunities, particularly the need for models that can integrate complex data to improve diagnostic accuracy and treatment outcomes.
Rafael Muñoz-Terol, Jesús Peral, Sandra Amador et al.· Heliyon· 4 citations· ⚡1
This paper identifies two different routes through which models can acquire geometrically separable features: they can learn them from complementary co-occurrence signals in general language data, including text-number co-occurrence and cross-number interaction, or from multi-token addition problems.
This work explores image generation using flow matching using flow matching and proposes an iterative process that can be integrated into virtually any generative modeling technique, thereby enhancing the performance and robustness of image synthesis systems.
Eldad Haber, Shadab Ahamed, Md Shahriar Rahim Siddiqui et al.· SIAM Journal on Scientific C...· 3 citations
This work presents AutoOR, a scalable synthetic data generation and reinforcement learning pipeline that trains LLMs to autoformulate optimization problems specified in natural language across linear, mixed-integer, and non-linear categories and introduces a curriculum RL strategy that bootstraps from limited initial training data to make this class tractable for post-training.
S. Motwani, Chuan Du, A. Petrov et al.· 2 citations
A benchmark built on the Speech Accessibility Project (SAP) dataset is introduced that tests whether diagnosis labels, clinician-derived speech ratings, and progressively richer clinical descriptions improve transcription accuracy for dysarthric speech, finding that current models do not meaningfully use this context.
P. Moure, Niclas Pokel, Bilal Bounajma et al.· arXiv.org· 2 citations
