• Turing Post
  • Posts
  • 🎄 Recap: 21 Episodes of AI 101 Guide

🎄 Recap: 21 Episodes of AI 101 Guide

Take some time to learn or refresh the key concepts, techniques, and models that matter most

Author

Ksenia Se
December 25, 2024

🎄 Merry Christmas, Happy Hanukkah and Joyful Kwanzaa to all who celebrates! 🎄

Hopefully, you have some time to unwind. But if unwinding for you means continuing to learn, we’ve got something for you: all our 101 AI guides published so far.

Hurry up, only now you can get full access to this collection with 40% OFF. You can also make it as a gift to someone else →

We appreciate you 🤍🩶

As a reminder, you can now find Turing Post on Hugging Face, and we just published there the article “🦸🏻#1: Open-endedness and AI Agents – A Path from Generative to Creative AI?”. You can read it there in full.

Now, dive in into AI 101 Series →

Topic 1: What is Mixture-of-Experts (MoE)?

we discuss the origins of MoE, why is it better than one neural network, Sparsely-Gated MoE, and sudden hype. Enjoy the collection of helpful links

www.turingpost.com/p/moe

Topic 2: What is Mamba?

We discuss sequence modeling, the drawbacks of transformers, and what Mamba brings into play

www.turingpost.com/p/mamba

Topic 3: What is Graph RAG approach?

we discuss the limitations of RAG and explore the benefits of Graph RAG approach, as well as clarify terms and provide a list of resources

www.turingpost.com/p/graphrag

Topic 4: What is Joint Embedding Predictive Architecture (JEPA)?

we discuss the Joint Embedding Predictive Architecture (JEPA), how it differs from transformers and provide you with list of models based on JEPA

www.turingpost.com/p/jepa

Topic 5: What is FSDP and YaFSDP?

we discuss the breakthroughs in GPU optimization techniques, exploring how YaFSDP surpasses FSDP in enhancing efficiency and scalability for large language mode

www.turingpost.com/p/yafsdp

Topic 6: What is KAN

we discuss how Kolmogorov-Arnold Networks (KANs) are redefining neural network architectures and their advantages over traditional multilayer perceptrons

www.turingpost.com/p/kan

Topic 7: What is LongRAG framework?

we discuss the limitations of RAG with a long-context window and explore the intuition behind the LongRAG framework to address these limitations, along with a list of resources for further learning

www.turingpost.com/p/longrag

Topic 8: What is LSTM and xLSTM?

we review what we know about LSTM networks and explore their new promising development – xLSTM

www.turingpost.com/p/xlstm

Topic 9: What is Speculative RAG?

We explore the recent Speculative RAG idea, highlighting where it excels, and discuss the limitations of other RAG systems. Bonus: Resources

www.turingpost.com/p/speculativerag

Topic 10: Inside DeepSeek Models

We discuss the innovation suggested by the DeepSeek team, how it improves the models' performance, and dive into the architectures and implementation

www.turingpost.com/p/deepseek

Topic 11: What are Chains (!) of Knowledge

we compare three distinct approaches, all called Chain of Knowledge, and suggest how they can be combined for better reasoning

www.turingpost.com/p/cok

Topic 12: What is HybridRAG?

we discuss the innovative combination of VectorRAG and GraphRAG in HybridRAG, its impact on financial document analysis and other areas of implementation, and clarify related terms for better understanding

www.turingpost.com/p/hybridrag

Topic 13: What is OLMoE?

Explore how open-source idea enhances MoE architecture

www.turingpost.com/p/olmoe

Topic 14: What are DoRA, QLoRA and QDoRA?

we compare three fine-tuning methods – DoRA, QLoRA, and QDoRA – each designed to improve model performance and memory efficiency in different ways.

www.turingpost.com/p/doraqloraqdora

Topic 15: Inside Whisper, an open-source audio model

Explore how OpenAI made their automatic speech recognition (ASR) model multilingual and multitasking

www.turingpost.com/p/topic-15-inside-whisper-an-open-source-audio-model

Topic 16: What is Whiteboard-of-Thought?

Explore how MLLMs can visually "think" step-by-step

www.turingpost.com/p/whiteboard-of-thought

Topic 17: Inside Les Ministraux

we trace Mistral's strategic roadmap and and unpack the unique performance of les Ministraux (Ministral)

www.turingpost.com/p/ministral

Topic 18: What is Mixture-of-Depths?

Explore how transformers can select different depths of processing and reduce compute needs

www.turingpost.com/p/mod

Topic 19: Inside LLaVA-o1

Let's explore a smarter Vision-Language Model (VLM) that thinks step-by-step

www.turingpost.com/p/llava-o1

Topic 20: What is Flow Matching?

Explore the key concepts of Flow Matching, its relation to diffusion models, and how it can enhance the training of generative models

www.turingpost.com/p/flowmatching

Topic 21: What is Natural Language Reinforcement Learning (NLRL)?

Explore how RL can be blended with natural language

www.turingpost.com/p/nlrl

Reply

or to participate.