How I Study AI - Learn AI Papers & Lectures the Easy Way

TAG-MoE: Task-Aware Gating for Unified Generative Mixture-of-Experts

Intermediate

Yu Xu, Hongbin Yan et al.Jan 12arXiv

TAG-MoE is a new way to steer Mixture-of-Experts (MoE) models using clear task hints, so the right “mini-experts” handle the right parts of an image job.

#Task-Aware Gating#Mixture-of-Experts#Unified Image Generation

NextFlow: Unified Sequential Modeling Activates Multimodal Understanding and Generation

Intermediate

Huichao Zhang, Liao Qu et al.Jan 5arXiv

NextFlow is a single, decoder-only Transformer that can read and write both text and images in one continuous sequence.

#Next-Scale Prediction#Autoregressive Transformer#Dual-Codebook Tokenization

Sparse-LaViDa: Sparse Multimodal Discrete Diffusion Language Models

Beginner

Shufan Li, Jiuxiang Gu et al.Dec 16arXiv

Sparse-LaViDa makes diffusion-style AI models much faster by skipping unhelpful masked tokens during generation while keeping quality the same.

#Masked Discrete Diffusion#Sparse Parameterization#Register Tokens

Papers3

TAG-MoE: Task-Aware Gating for Unified Generative Mixture-of-Experts

NextFlow: Unified Sequential Modeling Activates Multimodal Understanding and Generation

Sparse-LaViDa: Sparse Multimodal Discrete Diffusion Language Models