Best learning resources for AI
This paper proposes the Variational Lossy Autoencoder (VLAE), a VAE that uses autoregressive priors and decoders to deliberately discard local detail while retaining global structure. By limiting the receptive field of the PixelCNN decoder and employing autoregressive flows as the prior, the model forces the latent code to capture only high-level information, yielding controllable lossy representations. Experiments on MNIST, Omniglot, Caltech-101 Silhouettes and CIFAR-10 set new likelihood records for VAEs and demonstrate faithful global reconstructions with replaced textures. VLAE influenced research on representation bottlenecks, pixel-VAE hybrids, and state-of-the-art compression and generation benchmarks.
The book provides a comprehensive introduction to deep learning, covering foundational concepts like neural networks, optimization, convolutional and recurrent architectures, and probabilistic approaches. It bridges theory and practice, making it essential for both researchers and practitioners. Its impact has been profound, shaping modern AI research and education, inspiring breakthroughs in computer vision, natural language processing, and reinforcement learning, and serving as the go-to reference for anyone entering the deep learning field.
A SaaS-first MLOps suite that tracks experiments, datasets and models while enabling collaborative LLM/GenAI application development.
This paper introduces Message Passing Neural Networks (MPNNs), a unifying framework for graph-based deep learning, and applies it to quantum-chemistry property prediction, achieving state-of-the-art accuracy on the QM9 benchmark and approaching chemical accuracy on most targets. Its impact includes popularising graph neural networks, influencing subsequent work in cheminformatics, materials discovery, and the broader machine-learning community by demonstrating how learned message passing can replace hand-engineered molecular descriptors.
This paper introduces Relation Networks, a plug-and-play neural module that explicitly computes pair-wise object relations. When appended to standard CNN/LSTM encoders the module yields super-human 95.5 % accuracy on CLEVR, solves 18/20 bAbI tasks, and infers hidden links in dynamic physical systems, inspiring later work on relational reasoning across vision, language and RL.
The paper “Attention Is All You Need” (2017) introduced the Transformer — a novel neural architecture relying solely on self-attention, removing recurrence and convolutions. It revolutionized machine translation by dramatically improving training speed and translation quality (e.g., achieving 28.4 BLEU on English-German tasks), setting new state-of-the-art benchmarks. Its modular, parallelizable design opened the door to large-scale pretraining and fine-tuning, ultimately laying the foundation for modern large language models like BERT and GPT. This paper reshaped the landscape of NLP and deep learning, making attention-based models the dominant paradigm across many tasks.
Ray is an open-source distributed compute engine that lets you scale Python and AI workloads—from data processing to model training and serving—without deep distributed-systems expertise.
CNCF-incubating model inference platform (formerly KFServing) that provides Kubernetes CRDs for scalable predictive and generative workloads.
Open-source, high-performance server for deploying and scaling AI/ML models on GPUs or CPUs, supporting multiple frameworks and cloud/edge targets.
This paper introduces a Relational Memory Core that embeds multi-head dot-product attention into recurrent memory to enable explicit relational reasoning. Evaluated on synthetic distance-sorting, program execution, partially-observable reinforcement learning and large-scale language-modeling benchmarks, it consistently outperforms LSTM and memory-augmented baselines, setting state-of-the-art results on WikiText-103, Project Gutenberg and GigaWord. By letting memories interact rather than merely store information, the approach substantially boosts sequential relational reasoning and downstream task performance.
An open-source platform from Databricks that manages the entire machine-learning lifecycle with experiment tracking, model packaging, registry and deployment.
The BERT (Bidirectional Encoder Representations from Transformers) paper introduced a powerful pre-trained language model that uses deep bidirectional transformers and masked language modeling to capture both left and right context. Unlike prior unidirectional models, BERT achieved state-of-the-art performance across 11 NLP tasks (like GLUE, SQuAD) by enabling fine-tuning with minimal task-specific adjustments. Its impact reshaped NLP by setting a new standard for transfer learning, greatly improving accuracy on tasks such as question answering, sentiment analysis, and natural language inference, and inspiring a wave of follow-up models like RoBERTa, ALBERT, and T5.
