The Mysterious Math Behind LLMs

Dr Brian Keating
01:06:24 Report Issue
Loading transcript... Click for full transcript

Chapters & Sections (118)

00:00 Mathematical Foundations of Machine Learning chapter 2
00:00 Mathematical Foundations of Large Language Models
00:54 Mathematical Foundations of Machine Learning
01:48 Mathematical Foundations of Neural Networks chapter 2
01:48 Mathematical Foundations of Neural Networks
02:20 Mathematical Foundations of Machine Learning Algorithms
03:30 Early History of Artificial Neural Networks chapter 2
03:30 History of Artificial Neural Networks
04:29 Mathematical Foundations of Machine Learning
05:23 History of Machine Learning Development chapter 3
05:23 Mathematical Foundations of Machine Learning
06:27 Neural Network Architecture and Computation
06:53 Limitations of Early Neural Network Training Algorithms
07:26 Limitations of Single-Layer Neural Networks chapter 2
07:26 Limitations of Single Layer Neural Networks
08:16 Limitations of Early Neural Network Research
09:06 History of Neural Network Development chapter 2
09:06 History of Machine Learning Techniques
09:50 Early Neural Network Training Challenges
10:54 History of AI and Machine Learning Development chapter 2
10:54 History of AI and Machine Learning Development
12:10 GPU Matrix Manipulations in Machine Learning
12:40 The Evolution of AI Computing Infrastructure chapter 3
12:40 The Evolution of AI Computing Infrastructure
13:09 Training AI Models with Expert Feedback
13:39 Breaking Free from Unpaid Internships
13:59 High-Paying Opportunities in AI Research chapter 2
13:59 Benefits of Working with AI Liners
14:36 Mathematics in Machine Learning Models
15:24 The Phenomenon of Lockin in Technology chapter 3
15:24 The Dominance of Early Technology in AI
15:57 Optimization Trade-Offs in Design and Technology
16:48 Historical Influence on Infrastructure Design
17:22 Risk of Technological Lock-in with LLMs and GPUs chapter 2
17:22 Potential Lock-in with LLMs and GPUs
18:06 Risks of LLM Dominance in AI Landscape
19:02 Differences Between Human and Machine Learning chapter 1
19:02 Limitations of Top-Down AI Model Development
20:54 Quantum Gravity and AI Creativity Limitations chapter 2
20:54 Theoretical Foundations of Large Language Models
21:49 The Role of Embodiment in Human Intelligence
22:29 Materialism and the Limits of Computational Intelligence chapter 3
22:29 Materialism and Computational Principles of Consciousness
23:24 The Nature of Human Intelligence and Learning
23:56 The Nature of Consciousness and Intelligence
24:26 Consciousness and Machine Intelligence Limitations chapter 3
24:26 Consciousness in Artificial Intelligence Systems
25:08 Consciousness and Breakthroughs in AI
25:35 Mathematical Aspects of Machine Learning
26:11 Mathematical Spaces Behind Machine Learning Algorithms chapter 1
26:11 Mathematical Foundations of Large Language Models
27:50 Limitations of Large Language Model Data Saturation chapter 2
27:50 Limitations of Large Language Model Data Ingestion
28:57 Limitations of Large Language Model Advantages
29:37 Unlocking High-Quality Data for LLMs chapter 2
29:37 Unlocking High-Quality Data for LLMs
30:27 Risks of Overfitting in Large Language Models
31:09 Limitations of Current LLM Data Utilization chapter 2
31:09 Challenges of Using Private Data in LLMs
32:01 Efficiency of Human Learning vs Machine Learning
32:56 Limitations of Current Machine Learning Algorithms chapter 3
32:56 Limitations of Frozen Machine Learning Models
33:40 Alternatives to GPU-based LLMs
34:19 Hardware Requirements for Machine Learning Models
34:50 Alternative Approaches to Large Language Models chapter 2
34:50 Math Behind Large Language Model Architectures
35:47 Energy Efficiency in LLM Hardware Implementations
36:32 Energy Efficiency in Artificial Neural Networks chapter 1
36:32 Energy Efficiency in Artificial Neural Networks
38:10 Building Abstract World Models in AI Systems chapter 2
38:10 Constructing World Models in Machine Learning
39:17 Error Processing in LLMs and Human Brain
39:52 Data Efficiency in Machine Learning Models chapter 3
39:52 Internal Simulations in Machine Learning Models
40:30 Judging Books by Their Covers
40:59 Math Behind Machine Learning Algorithms
42:06 Mathematical Foundations of Large Language Models chapter 2
42:06 Mathematics Behind Large Language Models
42:42 Math Behind Large Language Models Explained
43:57 Overparameterization in Deep Learning Models Explained chapter 3
43:57 Overfitting vs. Large Model Parameters
44:30 Overparameterization in Machine Learning Models
45:00 Overfitting in Machine Learning Models
45:26 Implicit Regularization in Deep Neural Networks chapter 2
45:26 Implicit Regularization in Deep Neural Networks
46:44 Overparameterization in Machine Learning Methods
47:14 Math Behind Stochastic Gradient Descent Explained chapter
48:57 Math Behind Loss Landscapes in LLMs chapter 3
48:57 Math Behind Large Language Model Performance
49:35 Complexity of Loss Landscape in LLMs
50:14 Local Minima in Gradient Descent Optimization
50:55 Escaping Local Minima in Complex Optimization chapter 2
50:55 Escaping Local Minima in Machine Learning
51:27 Mathematical Foundations of Large Language Models
52:32 Artificial Neural Networks and Perceptron Basics chapter 2
52:32 Artificial Neural Networks and Perceptron Basics
53:39 Perceptron Algorithm and Image Classification
54:33 Mathematical Representation of Images in Neural Networks chapter 2
54:33 Mathematical Representation of Images in LLMs
55:56 Neural Network Architecture and Complexity
56:38 Neural Networks and Their Future Development chapter 3
56:38 Neural Network Architecture and Future Developments
57:25 Mathematical Foundations of Large Language Models
58:11 Predictive Power of Mathematical Equations
58:39 Mathematical Limitations of LLMs Predicted chapter 3
58:39 Pernicious Phenomena in Large Language Models
59:29 Probabilistic Nature of Next Token Prediction
01:00:02 Mathematical Limitations of Large Language Models
01:00:31 Mathematical Limitations of Large Language Models chapter 2
01:00:31 Mathematical Limitations of LLMs and Hallucinations
01:01:16 Emerging Phenomena in Complex Human Systems
01:02:23 Computational Nature of Human Perception chapter 2
01:02:23 Brain's Predictive Models of Perception and Reality
01:03:28 Psychosis and Machine Learning Similarities
01:04:18 Machine Hallucinations and Predictive Mechanisms chapter 3
01:04:18 Machine Hallucinations and Predictive Mechanisms
01:05:17 The Mysterious Math Behind Large Language Models
01:05:56 The Mysterious Math Behind Large Language Models

Transcript

Loading transcript...