Yannic Kilcher Courses

Author Interview - SayCan - Do As I Can, Not As I Say - Grounding Language in Robotic Affordances

Explore SayCan, a groundbreaking approach combining language models with robotic skills for complex task execution. Learn about its development, challenges, and future implications in AI and robotics.

• YouTube
• 59 minutes
• Self-Paced
• Free Video

Do As I Can, Not As I Say - Grounding Language in Robotic Affordances

Explanation of SayCan, a system combining language models with robotic skills for real-world task execution. Covers methodology, implementation, and experimental results for complex instruction following.

• YouTube
• 29 minutes
• Self-Paced
• Free Video

Author Interview - ACCEL- Evolving Curricula with Regret-Based Environment Design

Explore ACCEL, a novel approach combining regret-based sampling and level-editing for training robust AI agents. Learn about curriculum generation, minimax regret, and the future of reinforcement learning.

• YouTube
• 58 minutes
• Self-Paced
• Free Video

ACCEL - Evolving Curricula with Regret-Based Environment Design

Explore ACCEL, a novel approach combining regret-based sampling and level-editing for creating adaptive curricula in reinforcement learning, enhancing agent capabilities and robustness.

• YouTube
• 44 minutes
• Self-Paced
• Free Video

LAION-5B - 5 Billion Image-Text-Pairs Dataset - With the Authors

Explore LAION-5B, a massive open dataset of 5 billion image-text pairs. Learn about its creation, challenges, applications, and implications for AI research and development.

• YouTube
• 58 minutes
• Self-Paced
• Free Video

Author Interview - Transformer Memory as a Differentiable Search Index

Explore a novel approach to information retrieval using a single Transformer model that encodes corpus information in its parameters, enabling direct query-to-document mapping without external indices.

• YouTube
• 43 minutes
• Self-Paced
• Free Video

Transformer Memory as a Differentiable Search Index - Machine Learning Research Paper Explained

Explore a novel approach to information retrieval using a single Transformer model that encodes corpus information in its parameters, enabling direct query-to-document mapping without external indices.

• YouTube
• 52 minutes
• Self-Paced
• Free Video

The Weird and Wonderful World of AI Art

Explore the evolution of AI art, from early techniques to modern tools like CLIP and DALL-E. Discover the impact on artists, researchers, and amateurs, and ponder the future of this rapidly advancing field.

• YouTube
• 1 hour
• Self-Paced
• Free Video

Improving Intrinsic Exploration with Language Abstractions - Author Interview

Interview exploring how language abstractions can improve intrinsic exploration in reinforcement learning, discussing experimental results, challenges, and future directions in this promising research area.

• YouTube
• 49 minutes
• Self-Paced
• Free Video

Improving Intrinsic Exploration with Language Abstractions - Machine Learning Paper Explained

Explores using natural language to improve reinforcement learning agents' exploration in sparse-reward environments, demonstrating significant performance gains across challenging tasks in MiniGrid and MiniHack.

• YouTube
• 42 minutes
• Self-Paced
• Free Video

Author Interview - Memory-Assisted Prompt Editing to Improve GPT-3 After Deployment

Explore a method to improve GPT-3's performance post-deployment by maintaining interaction memory and dynamically adapting prompts, enhancing accuracy without retraining the model.

• YouTube
• 41 minutes
• Self-Paced
• Free Video

Memory-Assisted Prompt Editing to Improve GPT-3 After Deployment - Machine Learning Paper Explained

Explore adaptive methods to enhance GPT-3's performance post-deployment using memory-based prompts, improving accuracy without retraining the model. Learn about applications in fine-tuning and personalization.

• YouTube
• 37 minutes
• Self-Paced
• Free Video

Author Interview - Typical Decoding for Natural Language Generation

Interview exploring typical sampling, a new decoding method for language models that balances generating high-probability and high-information samples, aiming to produce more diverse and interesting text while reducing repetitive outputs.

• YouTube
• 49 minutes
• Self-Paced
• Free Video

BLIP- Bootstrapping Language-Image Pre-Training for Unified Vision-Language Understanding and Generation

BLIP: A versatile vision-language pre-training framework that excels in both understanding and generation tasks. Utilizes bootstrapping to improve noisy web data, achieving state-of-the-art results across various vision-language tasks.

• YouTube
• 47 minutes
• Self-Paced
• Free Video

Active Dendrites Avoid Catastrophic Forgetting - Interview With the Authors

Interview exploring how active dendrites in neural networks can mitigate catastrophic forgetting, enabling better multi-task and continual learning for AI systems in dynamic environments.

• YouTube
• 57 minutes
• Self-Paced
• Free Video