Tag: Mixture of Experts

A new commentary in Frontiers of Digital Education argues that open-weight large language models such as DeepSeek could extend personalized, ...

Researchers in China have developed a transparent deep learning framework called MDANet that accurately predicts the dynamic behavior of a ...

A new comprehensive review analyzes 13 pretrained multimodal AI models from 2020 to 2025, comparing their architectures, datasets, and applications ...

A new systematic review in Neural Computing and Applications maps the capabilities of modern large language models into six categories ...

Researchers in Beijing have developed a two-stage hybrid algorithm combining group relative policy optimization with Tabu search that reschedules disrupted ...

Researchers have created a hybrid onboard model for turboshaft engines that combines component-level physics simulations with sparse flight data calibration, ...

Researchers at Shanxi University have developed MAST-LLM, a framework that repurposes large language models with mixture-of-experts alignment and bidirectional spatio-temporal ...

A new diffusion-based framework called AffectMoE uses continuous valence–arousal soft routing over emotion-specific experts to generate 3D dance movements that ...

A new mixture-of-experts vision-language model called RSFG-MoE is designed to deliver fine-grained, high-resolution interpretation of remote sensing imagery.

A USC-developed blood-based screening algorithm cut the rate of unnecessary PET scans among Alzheimer's clinical trial candidates from over 70 ...