🎓How I Study AIHISA

📖Read

📄Papers 📰Blogs 🎬Courses

💡Learn

🛤️Paths 📚Topics 💡Concepts 🎴Shorts

🎯Practice

🧩Problems 🎯Prompts 🧠Review

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

Papers2

All Beginner Intermediate Advanced

All Sources arXiv

#Quantization

Rethinking Generative Recommender Tokenizer: Recsys-Native Encoding and Semantic Quantization Beyond LLMs

Yu Liang, Zhongjin Zhang et al.Feb 2arXiv

This paper proposes ReSID, a new way to turn items into short token codes (Semantic IDs) that are much easier for a recommender to predict.

#Semantic IDs#Generative Recommendation#Representation Learning

SnapGen++: Unleashing Diffusion Transformers for Efficient High-Fidelity Image Generation on Edge Devices

Dongting Hu, Aarush Gupta et al.Jan 13arXiv

This paper shows how to make powerful image‑generating Transformers run fast on phones without needing the cloud.

#Diffusion Transformer#Sparse Attention#Adaptive Sparse Self-Attention