Yannic Kilcher Courses

TransGAN - Two Transformers Can Make One Strong GAN - Machine Learning Research Paper Explained

Explore TransGAN, a novel approach using transformers for image generation. Learn about its architecture, key techniques, and competitive performance in various datasets.

• YouTube
• 30 minutes
• On-Demand
• Free Video

NFNets - High-Performance Large-Scale Image Recognition Without Normalization

Explore Normalizer-Free Networks, a novel approach to image recognition that achieves state-of-the-art accuracy without batch normalization, offering faster training and improved transfer learning performance.

• YouTube
• 34 minutes
• On-Demand
• Free Video

Nyströmformer- A Nyström-Based Algorithm for Approximating Self-Attention

Explore Nyströmformer, a novel algorithm for approximating self-attention in Transformers with linear complexity, enabling processing of longer sequences and improving scalability in natural language tasks.

• YouTube
• 48 minutes
• On-Demand
• Free Video

Feedback Transformers - Addressing Some Limitations of Transformers with Feedback Memory

Explores Feedback Transformers, an architecture that improves on standard Transformers by allowing higher-level representations to inform lower-level processing, enhancing performance on complex reasoning and long-range dependency tasks.

• YouTube
• 44 minutes
• On-Demand
• Free Video

SingularityNET - A Decentralized, Open Market and Network for AIs

Explore SingularityNET's decentralized AI marketplace, enabling open development and interoperability of AI services. Learn about its blockchain foundation, token economy, and potential impact on the future of artificial intelligence.

• YouTube
• 52 minutes
• On-Demand
• Free Video

Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity

Explore Switch Transformers: a novel approach to scaling AI models to trillion parameters using sparsity and hard routing, offering improved speed and efficiency compared to dense counterparts.

• YouTube
• 34 minutes
• On-Demand
• Free Video

OpenAI CLIP- Connecting Text and Images

Detailed explanation of OpenAI's CLIP model, which connects text and images using contrastive learning on 400M image-text pairs, enabling zero-shot transfer to various computer vision tasks.

• YouTube
• 48 minutes
• On-Demand
• Free Video

OpenAI DALL·E - Creating Images from Text - Blog Post Explained

Explore OpenAI's DALL·E model, its architecture, capabilities, and limitations in generating high-quality images from text descriptions. Learn about its potential impact on AI applications combining text and visuals.

• YouTube
• 56 minutes
• On-Demand
• Free Video

ReBeL - Combining Deep Reinforcement Learning and Search for Imperfect-Information Games

Explores ReBeL, an AI framework combining reinforcement learning and search for imperfect-information games. Explains its application to poker, convergence to Nash equilibrium, and superhuman performance with minimal domain knowledge.

• YouTube
• 1 hour 12 minutes
• On-Demand
• Free Video

Predictive Coding Approximates Backprop Along Arbitrary Computation Graphs - Paper Explained

Explores how predictive coding can approximate backpropagation in neural networks, bridging neuroscience and machine learning. Discusses implications for biologically plausible deep learning algorithms.

• YouTube
• 48 minutes
• On-Demand
• Free Video

Fourier Neural Operator for Parametric Partial Differential Equations - Paper Explained

Explore a novel neural network architecture for solving partial differential equations efficiently. Learn about Fourier Neural Operators and their potential to revolutionize scientific computing and engineering simulations.

• YouTube
• 1 hour 6 minutes
• On-Demand
• Free Video

Underspecification Presents Challenges for Credibility in Modern Machine Learning - Paper Explained

Explores underspecification in ML pipelines, where models with similar training performance behave differently in real-world deployment, leading to instability and poor outcomes across various domains.

• YouTube
• 59 minutes
• On-Demand
• Free Video

Language Models Are Open Knowledge Graphs

Explore how pre-trained language models can be used to construct knowledge graphs without human supervision, uncovering new relations and outperforming existing methods in automated KG construction.

• YouTube
• 52 minutes
• On-Demand
• Free Video

Rethinking Attention with Performers

Explore a novel approach to efficient attention mechanisms in Transformers using random positive orthogonal features, enabling linear-time approximation of attention matrices for scalable deep learning architectures.

• YouTube
• 55 minutes
• On-Demand
• Free Video

LambdaNetworks- Modeling Long-Range Interactions Without Attention

Explore LambdaNetworks, a novel approach to modeling long-range interactions in AI without traditional attention mechanisms, offering improved efficiency and performance in computer vision tasks.

• YouTube
• 1 hour
• On-Demand
• Free Video