songqiang321/Awesome-AI-Papers

Organized collection of 500+ foundational and recent papers with accompanying implementations across NLP (Word2Vec to LLMs), Computer Vision, multimodal learning, reinforcement learning, and graph neural networks, structured by sub-domains like LLM techniques (RAG, prompt learning, MoE), applications (agents, code generation, medical/financial), and specialized topics (alignment, hallucination mitigation). Curates research spanning from classical architectures (Transformers, Seq2Seq) through state-of-the-art models (GPT series, BERT variants, diffusion models, multimodal LLMs) with direct links to both papers and official/community implementations. Serves as a taxonomy-driven learning resource for practitioners tracking the evolution of deep learning, particularly suited for understanding LLM development trajectories and contemporary applications in specialized domains.