正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Functions are vectors! This perspective lets us apply the tools of linear algebra to computational problems from image and geometry processing to machine learning and light transport—and provides a natural explanation for Fourier series. Let's explore:

Max Slater

2,957 subscribers

30,424 次观看 • 3 年前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

🧵1/n This thread covers the linear algebra for modern computer graphics, bridging geometry, physics, and computation. Video 1: Vector Spaces & Cartesian Coordinates Source: CMU. 👇Video 2: Vector Operations, Vector Spaces, Functions as Vectors, and Vectors in Coordinates

🧵1/n This thread covers the linear algebra for modern computer graphics, bridging geometry, physics, and computation. Video 1: Vector Spaces & Cartesian Coordinates Source: CMU. 👇Video 2: Vector Operations, Vector Spaces, Functions as Vectors, and Vectors in Coordinates

tetsuo

52,043 次观看 • 1 年前

[CLIP] by Hand ✍️ The CLIP (Contrastive Language–Image Pre-training) model, a groundbreaking work by OpenAI, redefines the intersection of computer vision and natural language processing. It is the basis of all the multi-modal foundation models we see today. How does CLIP work? Goal: 🟨 Learn a shared embedding space for text and image [1] Given ↳ A mini batch of 3 text-image pairs ↳ OpenAI used 400 million text-image pairs to train its original CLIP model. Process 1st pair: "big table" [2] 🟪 Text → 2 Vectors (3D) ↳ Look up word embedding vectors using word2vec. [3] 🟩 Image → 2 Vectors (4D) ↳ Divide the image into two patches. ↳ Flatten each patch [4] Process other pairs ↳ Repeat [2]-[3] [5] 🟪 Text Encoder & 🟩 Image Encoder ↳ Encode input vectors into feature vectors ↳ Here, both encoders are simple one layer perceptron (linear + ReLU) ↳ In practice, the encoders are usually transformer models. [6] 🟪 🟩 Mean Pooling: 2 → 1 vector ↳ Average 2 feature vectors into a single vector by averaging across the columns ↳ The goal is to have one vector to represent each image or text [7] 🟪 🟩 -> 🟨 Projection ↳ Note that the text and image feature vectors from the encoders have different dimensions (3D vs. 4D). ↳ Use a linear layer to project image and text vectors to a 2D shared embedding space. 🏋️ Contrastive Pre-training 🏋️ [8] Prepare for MatMul ↳ Copy text vectors (T1,T2,T3) ↳ Copy the transpose of image vectors (I1,I2,I3) ↳ They are all in the 2D shared embedding space. [9] 🟦 MatMul ↳ Multiply T and I matrices. ↳ This is equivalent to taking dot product between every pair of image and text vectors. ↳ The purpose is to use dot product to estimate the similarity between a pair of image-text. [10] 🟦 Softmax: e^x ↳ Raise e to the power of the number in each cell ↳ To simplify hand calculation, we approximate e^□ with 3^□. [11] 🟦 Softmax: ∑ ↳ Sum each row for 🟩 image→🟪 text ↳ Sum each column for 🟪 text→ 🟩 image [12] 🟦 Softmax: 1 / sum ↳ Divide each element by the column sum to obtain a similarity matrix for 🟪 text→🟩 image ↳ Divide each element by the row sum to obtain a similarity matrix for 🟩 image→🟪 text [13] 🟥 Loss Gradients ↳ The "Targets" for the similarity matrices are Identity Matrices. ↳ Why? If I and T come from the same pair (i=j), we want the highest value, which is 1, and 0 otherwise. ↳ Apply the simple equation of [Similarity - Target] to compute gradients of for both directions. ↳ Why so simple? Because when Softmax and Cross-Entropy Loss are used together, the math magically works out that way. ↳ These gradients kick off the backpropagation process to update weights and biases of the encoders and projection layers (red borders).

[CLIP] by Hand ✍️ The CLIP (Contrastive Language–Image Pre-training) model, a groundbreaking work by OpenAI, redefines the intersection of computer vision and natural language processing. It is the basis of all the multi-modal foundation models we see today. How does CLIP work? Goal: 🟨 Learn a shared embedding space for text and image [1] Given ↳ A mini batch of 3 text-image pairs ↳ OpenAI used 400 million text-image pairs to train its original CLIP model. Process 1st pair: "big table" [2] 🟪 Text → 2 Vectors (3D) ↳ Look up word embedding vectors using word2vec. [3] 🟩 Image → 2 Vectors (4D) ↳ Divide the image into two patches. ↳ Flatten each patch [4] Process other pairs ↳ Repeat [2]-[3] [5] 🟪 Text Encoder & 🟩 Image Encoder ↳ Encode input vectors into feature vectors ↳ Here, both encoders are simple one layer perceptron (linear + ReLU) ↳ In practice, the encoders are usually transformer models. [6] 🟪 🟩 Mean Pooling: 2 → 1 vector ↳ Average 2 feature vectors into a single vector by averaging across the columns ↳ The goal is to have one vector to represent each image or text [7] 🟪 🟩 -> 🟨 Projection ↳ Note that the text and image feature vectors from the encoders have different dimensions (3D vs. 4D). ↳ Use a linear layer to project image and text vectors to a 2D shared embedding space. 🏋️ Contrastive Pre-training 🏋️ [8] Prepare for MatMul ↳ Copy text vectors (T1,T2,T3) ↳ Copy the transpose of image vectors (I1,I2,I3) ↳ They are all in the 2D shared embedding space. [9] 🟦 MatMul ↳ Multiply T and I matrices. ↳ This is equivalent to taking dot product between every pair of image and text vectors. ↳ The purpose is to use dot product to estimate the similarity between a pair of image-text. [10] 🟦 Softmax: e^x ↳ Raise e to the power of the number in each cell ↳ To simplify hand calculation, we approximate e^□ with 3^□. [11] 🟦 Softmax: ∑ ↳ Sum each row for 🟩 image→🟪 text ↳ Sum each column for 🟪 text→ 🟩 image [12] 🟦 Softmax: 1 / sum ↳ Divide each element by the column sum to obtain a similarity matrix for 🟪 text→🟩 image ↳ Divide each element by the row sum to obtain a similarity matrix for 🟩 image→🟪 text [13] 🟥 Loss Gradients ↳ The "Targets" for the similarity matrices are Identity Matrices. ↳ Why? If I and T come from the same pair (i=j), we want the highest value, which is 1, and 0 otherwise. ↳ Apply the simple equation of [Similarity - Target] to compute gradients of for both directions. ↳ Why so simple? Because when Softmax and Cross-Entropy Loss are used together, the math magically works out that way. ↳ These gradients kick off the backpropagation process to update weights and biases of the encoders and projection layers (red borders).

Tom Yeh

67,858 次观看 • 2 年前

Dr. Robert Snellman's eight-hour course: The Fundamentals of Mathematics: Algebra, is available now. Enroll now for immediate access at Peterson Academy. Accompanying the course are practice questions as well! In this course, Dr. Snellman offers an accelerated and comprehensive introduction to algebra. He begins with basic variable manipulation and linear equations, then progresses to polynomials, quadratic equations, and exponential functions. Guided by Dr. Snellman, we build a strong mathematical foundation through a systematic study of functions, graphs, and solution methods, with a consistent emphasis on real-world applications in physics, finance, and cryptography. The course introduces increasingly sophisticated mathematical concepts and culminates in systems of linear equations and matrix operations, demonstrating how algebraic thinking can model and solve meaningful, practical problems.

Dr. Robert Snellman's eight-hour course: The Fundamentals of Mathematics: Algebra, is available now. Enroll now for immediate access at Peterson Academy. Accompanying the course are practice questions as well! In this course, Dr. Snellman offers an accelerated and comprehensive introduction to algebra. He begins with basic variable manipulation and linear equations, then progresses to polynomials, quadratic equations, and exponential functions. Guided by Dr. Snellman, we build a strong mathematical foundation through a systematic study of functions, graphs, and solution methods, with a consistent emphasis on real-world applications in physics, finance, and cryptography. The course introduces increasingly sophisticated mathematical concepts and culminates in systems of linear equations and matrix operations, demonstrating how algebraic thinking can model and solve meaningful, practical problems.

Peterson Academy

59,142 次观看 • 1 年前

Watch our co-founder, , explain how the future of machine learning will lead us to a continuously updated digital representation of the world on the internet, and why this requires shifts in the way we access and allocate computational resources.

Watch our co-founder, , explain how the future of machine learning will lead us to a continuously updated digital representation of the world on the internet, and why this requires shifts in the way we access and allocate computational resources.

gensyn

29,885 次观看 • 1 年前

Data preparation! It's crucial for machine learning, and we all hate it. Tools and techniques to reduce this burden? A quick summary of 10 years of R&D on this, from cheap tricks to LLMs and graph neural networks 1/9

Data preparation! It's crucial for machine learning, and we all hate it. Tools and techniques to reduce this burden? A quick summary of 10 years of R&D on this, from cheap tricks to LLMs and graph neural networks 1/9

Gael Varoquaux 🦋

13,757 次观看 • 1 年前

a playlist of 30 youtube videos to learn machine learning fundamentals from scratch if you're struggling on where to start learning ML, this list goes this "Machine Learning: Teach by Doing" is a solid choice to learn both theory and code. (1) Introduction to Machine Learning Teach by Doing: (2) What is Machine Learning? History of Machine Learning: (3) Types of ML Models: (4) 6 steps of any ML project: (5) Install Python and VSCode and run your first code: (6) Linear Classifiers Part 1: (7) Linear Classifiers Part 2: (8) Jupyter Notebook, Numpy and Scikit-Learn: (9) Running the Random Linear Classifier Algorithm in Python: (10) The oldest ML model - Perceptron: (11) Coding the Perceptron: (12) Perceptron Convergence Theorem: (13) Magic of features in Machine Learning: (14) One hot encoding: (15) Logistic Regression Part 1: (16) Cross Entropy Loss: (17) How gradient descent works: (18) Logistic Regression from scratch in Python: (19) Introduction to Regularization: (20) Implementing Regularization in Python: (21) Linear Regression Introduction: (22) Ordinary Least Squares step by step implementation: (23) Ridge regression fundamentals and intuition: (24) Regression recap for interviews: (25) Neural network architecture in 30 minutes: (26) Backpropagation intuition: (27) Neural network activation functions: (28) Momentum in gradient descent: (29) Hands on neural network training in Python: (30) Introduction to Convolutional Neural Networks (CNNs):

a playlist of 30 youtube videos to learn machine learning fundamentals from scratch if you're struggling on where to start learning ML, this list goes this "Machine Learning: Teach by Doing" is a solid choice to learn both theory and code. (1) Introduction to Machine Learning Teach by Doing: (2) What is Machine Learning? History of Machine Learning: (3) Types of ML Models: (4) 6 steps of any ML project: (5) Install Python and VSCode and run your first code: (6) Linear Classifiers Part 1: (7) Linear Classifiers Part 2: (8) Jupyter Notebook, Numpy and Scikit-Learn: (9) Running the Random Linear Classifier Algorithm in Python: (10) The oldest ML model - Perceptron: (11) Coding the Perceptron: (12) Perceptron Convergence Theorem: (13) Magic of features in Machine Learning: (14) One hot encoding: (15) Logistic Regression Part 1: (16) Cross Entropy Loss: (17) How gradient descent works: (18) Logistic Regression from scratch in Python: (19) Introduction to Regularization: (20) Implementing Regularization in Python: (21) Linear Regression Introduction: (22) Ordinary Least Squares step by step implementation: (23) Ridge regression fundamentals and intuition: (24) Regression recap for interviews: (25) Neural network architecture in 30 minutes: (26) Backpropagation intuition: (27) Neural network activation functions: (28) Momentum in gradient descent: (29) Hands on neural network training in Python: (30) Introduction to Convolutional Neural Networks (CNNs):

ℏεsam

117,570 次观看 • 1 年前

$TAO is tackling the core challenges in machine learning, from optimizing data collection to faster inference, aiming to outperform giants like OpenAI and Microsoft. With a focus on computational efficiency and robust models, the race is on.

$TAO is tackling the core challenges in machine learning, from optimizing data collection to faster inference, aiming to outperform giants like OpenAI and Microsoft. With a focus on computational efficiency and robust models, the race is on.

Grayscale

88,595 次观看 • 1 年前

New Book & Video Series!!! (late 2025) Optimization Bootcamp: Applications in Machine Learning, Control, and Inverse Problems Comment for a sneak peak to help proofread and I'll DM (proof reading, typos, HW problems, all get acknowledgment in book!)

New Book & Video Series!!! (late 2025) Optimization Bootcamp: Applications in Machine Learning, Control, and Inverse Problems Comment for a sneak peak to help proofread and I'll DM (proof reading, typos, HW problems, all get acknowledgment in book!)

Steven Brunton

103,238 次观看 • 1 年前

Early careers roles at Williams are still live. We are on a long-term mission to return to the front of the grid and are looking for the brightest and best talent to help us get there. Visit our careers page to explore opportunities and apply ⬇️

Early careers roles at Williams are still live. We are on a long-term mission to return to the front of the grid and are looking for the brightest and best talent to help us get there. Visit our careers page to explore opportunities and apply ⬇️

Atlassian Williams F1 Team

220,613 次观看 • 1 年前

Today we’re introducing SceneScript, a novel method for reconstructing environments and representing the layout of physical spaces from Reality Labs at Meta Research. Details ➡️ SceneScript is able to directly infer a room’s geometry using end-to-end machine learning and represent it using language. Compared to previous approaches, this results in representations of physical scenes that are compact, complete, interpretable and extensible.

Today we’re introducing SceneScript, a novel method for reconstructing environments and representing the layout of physical spaces from Reality Labs at Meta Research. Details ➡️ SceneScript is able to directly infer a room’s geometry using end-to-end machine learning and represent it using language. Compared to previous approaches, this results in representations of physical scenes that are compact, complete, interpretable and extensible.

AI at Meta

334,490 次观看 • 2 年前

if you're struggling on where to start learning ML, here’s a playlist of 30 youtube videos to learn machine learning fundamentals from scratch "Machine Learning: Teach by Doing" is a solid choice to learn both theory and code. (1) Introduction to Machine Learning Teach by Doing: (2) What is Machine Learning? History of Machine Learning: (3) Types of ML Models: (4) 6 steps of any ML project: (5) Install Python and VSCode and run your first code: (6) Linear Classifiers Part 1: (7) Linear Classifiers Part 2: (8) Jupyter Notebook, Numpy and Scikit-Learn: (9) Running the Random Linear Classifier Algorithm in Python: (10) The oldest ML model - Perceptron: (11) Coding the Perceptron: (12) Perceptron Convergence Theorem: (13) Magic of features in Machine Learning: (14) One hot encoding: (15) Logistic Regression Part 1: (16) Cross Entropy Loss: (17) How gradient descent works: (18) Logistic Regression from scratch in Python: (19) Introduction to Regularization: (20) Implementing Regularization in Python: (21) Linear Regression Introduction: (22) Ordinary Least Squares step by step implementation: (23) Ridge regression fundamentals and intuition: (24) Regression recap for interviews: (25) Neural network architecture in 30 minutes: (26) Backpropagation intuition: (27) Neural network activation functions: (28) Momentum in gradient descent: (29) Hands on neural network training in Python: (30) Introduction to Convolutional Neural Networks (CNNs):

if you're struggling on where to start learning ML, here’s a playlist of 30 youtube videos to learn machine learning fundamentals from scratch "Machine Learning: Teach by Doing" is a solid choice to learn both theory and code. (1) Introduction to Machine Learning Teach by Doing: (2) What is Machine Learning? History of Machine Learning: (3) Types of ML Models: (4) 6 steps of any ML project: (5) Install Python and VSCode and run your first code: (6) Linear Classifiers Part 1: (7) Linear Classifiers Part 2: (8) Jupyter Notebook, Numpy and Scikit-Learn: (9) Running the Random Linear Classifier Algorithm in Python: (10) The oldest ML model - Perceptron: (11) Coding the Perceptron: (12) Perceptron Convergence Theorem: (13) Magic of features in Machine Learning: (14) One hot encoding: (15) Logistic Regression Part 1: (16) Cross Entropy Loss: (17) How gradient descent works: (18) Logistic Regression from scratch in Python: (19) Introduction to Regularization: (20) Implementing Regularization in Python: (21) Linear Regression Introduction: (22) Ordinary Least Squares step by step implementation: (23) Ridge regression fundamentals and intuition: (24) Regression recap for interviews: (25) Neural network architecture in 30 minutes: (26) Backpropagation intuition: (27) Neural network activation functions: (28) Momentum in gradient descent: (29) Hands on neural network training in Python: (30) Introduction to Convolutional Neural Networks (CNNs):

ℏεsam

109,292 次观看 • 1 年前

The theory of higher order topological dynamics, which combines multilevel interactions between discrete topology and nonlinear dynamics, has the potential to enhance our understanding of complex systems such as the functions of the nervous system, the development of next-generation machine learning and the creation of advanced nodal processing algorithms. An important and unexpected collective behavior of signal processing in multilevel nodal networks has been observed to lead to a synchronization and diffusion of the irrotational and the solenoidal components of the systems revealing a deep relation of these mechanisms with the complexity of discrete topology. The perspective of the preliminary study linked here offers insights into how topology morphs dynamics, how dynamics stem from topology and how topology evolves dynamically. 🔗

The theory of higher order topological dynamics, which combines multilevel interactions between discrete topology and nonlinear dynamics, has the potential to enhance our understanding of complex systems such as the functions of the nervous system, the development of next-generation machine learning and the creation of advanced nodal processing algorithms. An important and unexpected collective behavior of signal processing in multilevel nodal networks has been observed to lead to a synchronization and diffusion of the irrotational and the solenoidal components of the systems revealing a deep relation of these mechanisms with the complexity of discrete topology. The perspective of the preliminary study linked here offers insights into how topology morphs dynamics, how dynamics stem from topology and how topology evolves dynamically. 🔗

Maurizio Iβλἄ

40,749 次观看 • 1 年前

ResNet by hand ✍️ ~ 10 steps walkthrough below "Deep Residual Learning for Image Recognition" (Kaiming He, CVPR 2016) is among the most cited papers in all of deep learning. Why does it matter so much? It fixed the exploding and vanishing gradients that kept deep networks from being deep, and made thousands of layers possible. How simple was the fix? An identity matrix. Goal: push three input vectors through a residual block, then through a transformer encoder block, filling in every cell yourself. = 1. Given = A mini batch of three input vectors, 3D, and the weights of the layers ahead. = 2. Linear layer = Let us multiply by the weights, add the bias, and apply ReLU so negatives become 0. Three feature vectors out. This is F(X). = 3. Concatenate = Now the trick. Stack an identity matrix beside the second layer's weights, and stack the input vectors under the features. Draw the lines between rows and columns: those are the skip connections. The identity is the residual. = 4. Linear layer + identity = We multiply the two stacked matrices. The identity carries X straight through while the weights transform it, so a single multiplication computes F(X) + X. Apply ReLU and hand it to the next block. Now watch the same trick inside a transformer, first in attention. = 5. Attention = Let us take three input vectors in 2D, compute the attention matrix, and multiply to get attention weighted vectors. = 6. Concatenate = We stack two identities this time, two residuals, which is how you get 1 + 1, and stack the input vectors with the attention weighted ones. = 7. Add = Multiply the stacked matrices. The identity adds attention to its own input, across the columns, which is how positions get combined. And again in the feed forward layer. = 8. First layer = Let us multiply by the feed forward weights and bias, then ReLU. Three feature vectors. = 9. Concatenate = Stack and link exactly as in step 3: the residual again. = 10. Second layer + identity = We multiply, apply ReLU, and pass the result to the next encoder block. This identity adds across the rows, combining features rather than positions. Takeaway: one simple "add" is what made really deep networks possible. 💾 Save this post!

ResNet by hand ✍️ ~ 10 steps walkthrough below "Deep Residual Learning for Image Recognition" (Kaiming He, CVPR 2016) is among the most cited papers in all of deep learning. Why does it matter so much? It fixed the exploding and vanishing gradients that kept deep networks from being deep, and made thousands of layers possible. How simple was the fix? An identity matrix. Goal: push three input vectors through a residual block, then through a transformer encoder block, filling in every cell yourself. = 1. Given = A mini batch of three input vectors, 3D, and the weights of the layers ahead. = 2. Linear layer = Let us multiply by the weights, add the bias, and apply ReLU so negatives become 0. Three feature vectors out. This is F(X). = 3. Concatenate = Now the trick. Stack an identity matrix beside the second layer's weights, and stack the input vectors under the features. Draw the lines between rows and columns: those are the skip connections. The identity is the residual. = 4. Linear layer + identity = We multiply the two stacked matrices. The identity carries X straight through while the weights transform it, so a single multiplication computes F(X) + X. Apply ReLU and hand it to the next block. Now watch the same trick inside a transformer, first in attention. = 5. Attention = Let us take three input vectors in 2D, compute the attention matrix, and multiply to get attention weighted vectors. = 6. Concatenate = We stack two identities this time, two residuals, which is how you get 1 + 1, and stack the input vectors with the attention weighted ones. = 7. Add = Multiply the stacked matrices. The identity adds attention to its own input, across the columns, which is how positions get combined. And again in the feed forward layer. = 8. First layer = Let us multiply by the feed forward weights and bias, then ReLU. Three feature vectors. = 9. Concatenate = Stack and link exactly as in step 3: the residual again. = 10. Second layer + identity = We multiply, apply ReLU, and pass the result to the next encoder block. This identity adds across the rows, combining features rather than positions. Takeaway: one simple "add" is what made really deep networks possible. 💾 Save this post!

Tom Yeh

17,988 次观看 • 17 天前

Quick demo of the voice to voice translation between English and Twi that you can try today on our web app. This one is from the perspective of a tourist visiting Ghana and needing to get around. Brought to you by the amazing team at Ghana Natural Language Processing (NLP) Algorine Research

Quick demo of the voice to voice translation between English and Twi that you can try today on our web app. This one is from the perspective of a tourist visiting Ghana and needing to get around. Brought to you by the amazing team at Ghana Natural Language Processing (NLP) Algorine Research

Paul Azunre

165,100 次观看 • 2 年前

Cubes are one of the most powerful tools for learning to draw, When you modify cubes you’re training perspective, form, and line control all at once. This is how artists learn to see objects in 3D instead of copying flat shapes.

Cubes are one of the most powerful tools for learning to draw, When you modify cubes you’re training perspective, form, and line control all at once. This is how artists learn to see objects in 3D instead of copying flat shapes.

ANGRYPOP ©

90,836 次观看 • 5 个月前

AAA is screwed. Power to the gamers. Google's Genie 3 makes a game from image or prompt, let's you run around and explore. Things are happening fast. Traditional ways of making games is going to disappear.

AAA is screwed. Power to the gamers. Google's Genie 3 makes a game from image or prompt, let's you run around and explore. Things are happening fast. Traditional ways of making games is going to disappear.

Grummz

142,463 次观看 • 6 个月前

Excited to launch FieldDay (FieldDay) today. It’s the culmination of years of work in working on approachable tools for machine learning with a data-centric approach, and a first step towards enabling SMEs and enthusiasts to build intelligent apps end to end.

Excited to launch FieldDay (FieldDay) today. It’s the culmination of years of work in working on approachable tools for machine learning with a data-centric approach, and a first step towards enabling SMEs and enthusiasts to build intelligent apps end to end.

Aaron Abentheuer

20,910 次观看 • 3 年前

Quick Take | How did Jawaharlal Nehru apply the Reconquista of Spain to shape India’s Hindu nationalism? Explore this historical perspective with Salman Javed and Syed Muhammad Ali.

Quick Take | How did Jawaharlal Nehru apply the Reconquista of Spain to shape India’s Hindu nationalism? Explore this historical perspective with Salman Javed and Syed Muhammad Ali.

Pakistan Now

35,903 次观看 • 1 年前