Laplacian eigenmaps perform non-linear dimensionality reduction ("manifold learning") by... embedding data points using the eigenvectors of a graph Laplacian as coordinates.show more

I wrote this fun interactive article explaining Isomap, a... classic non-linear dimensionality reduction technique. Isomap captures the local similarity structure of data through a graph and computes an embedding that preserves this structure. Check it out! Link below.show more

Many dimensionality reduction algorithms share a few central principles.... 1. Construct a graph that captures the data's local structure 2. Measure "geodesic" distances between points using the graph 3. Project the points to a lower dimension while preserving these distancesshow more

Can we make Transformers better and more efficient for... robot learning? Excited to introduce Body Transformer (BoT), an architecture that leverages robot embodiment in the attention mechanism, by treating it as a graph of sensors and actuators.show more

Footage, released by Hamas, of a Hamas terrorist using... a civilian area as a launching point. Here is the criteria for Hamas’ rocket launching points: Anywhere. Areas with civilians preferred.show more

#cellbiology #bioart The actin #cytoskeleton (green) is linked to... a variety of transmembrane proteins through alpha-actinin (magenta), with such linkages being particularly dense in membrane ruffles and filopodia. As seen at 3 frames/min for 8 min by linear and non-linear structured illumination microscopy (97 and 60 nm resolution, respectively).show more

Amazon’s machine learning model collects 300 million data points... per season and can now predict which players are likely to blitz before the snap. This is a look into the future of broadcasting.show more

USAAF airmen stationed in in Southern Italy learning a... few choice phrases in the local language from a young teacher using the side of a B-24 bomber as a blackboard in 1944show more

Non-binary actress Emma Corrin dresses like a bird in... bizarre feather outfit during Paris fashion show. The star of Netflix ‘The Crown’ identifies as queer non-binary using they/them pronouns.show more

How can we address the scarcity of data required... for specialized AI? Learn about Simula, a framework that reframes synthetic data generation as dataset-level mechanism design. By using reasoning to architect datasets from first principles, Simula enables fine-grained control over coverage, complexity, and quality. More →show more

Dry Needling: Ultrasound Diagnosis A patient presents with pain... along the semitendinosus muscle and a suspected muscle fiber tear. A thigh ultrasound is ordered. Ultrasound reveals multiple linear tracts of increased echogenicity in the proximal third of the semitendinosus muscle, which correlate with a history of dry needling performed 3 days before the radiological study. Dry needling is an invasive technique used in physiotherapy to eliminate painful muscle points (or trigger points) that can develop due to overuse. It is essential not to perform this treatment within 5 days of a diagnostic ultrasound or MRI, as it can hinder or prevent the diagnosis of a muscle injury, especially a grade I injury.show more

For the Godot Shaders Bible, we’re taking a different... approach: starting with simple shaders and learning as we go. Topics will include: - Profiling and optimization. - Compute shaders. - Particle systems. - Ray marching. - Maths (quaternions, matrices, etc.) - VFX. Everything will be presented in a linear, step-by-step way, just as always! Subscribe today! 👉 #GODOT #GodotEngine #IndieDevs #indiegameshow more

North Gaza City Today | Posted by Israeli non-military... individuals entering the strip. This was taken in north Gaza city near the water station they are using as a military base 1/2show more

Posted a full preview of this in my discord,... here's a snippet! This is the culmination of all the stuff I've been learning the last few months. There's a Daz3d model in there being driven by motion capture data, with a physics driven hair rig and some basic jiggle on the boobs.show more

Imagine: Parallel Neural Networks with Shaders 🤔 (GLSL instead... of CUDA) Parallelization by encoding the weights and activations as Buffer Objects or Textures, using fs to perform matmul and activation functions in parallel. Each fragm shader instance parproc 1..n neuron.show more

My data has been exhausted once again, but this... time, my mobile data isn’t sharing as expected since I don’t have a bundle. This means yesterday, my data bundle was being deducted at a rate of 10MB per second, even though my internet speed or the data my device was using wasn’t even up to 2MB per second. Inaccurate deductions!show more

🚨Edmonton: #Khalistani activists in Canada are openly radicalizing young... children by parading them in processions where kids perform with swords & chant separatist slogans. This is not cultural celebration — it is pure exploitation, using innocent children as props to push a violent separatist agenda under the guise of a 'Nagar Kirtan'. #cdnpolishow more

19 years ago today: DeMarco Murray committed to OU... over USC. A 5-star prospect out of Bishop Gorman in Las Vegas, Demarco ranked as the No. 22 overall player in the 2006 class. He lived up to the hype, and then some, finishing his career with 390 points, the most by a non-kicker in OU history.show more

As a learning test, I ported part of my... 2D Cloth Physics Simulation to the web using #Rust 🦀 🧪 Try the mini demo: Real-time simulation with Verlet integration, touch/mouse interaction, and web delivery. #gamedev #physics #simulation #verletintegrationshow more

Vector Database by Hand ✍️ Vector databases are revolutionizing... how we search and analyze complex data. They have become the backbone of Retrieval Augmented Generation (#RAG). How do vector databases work? [1] Given ↳ A dataset of three sentences, each has 3 words (or tokens) ↳ In practice, a dataset may contain millions or billions of sentences. The max number of tokens may be tens of thousands (e.g., 32,768 mistral-7b). Process "how are you" [2] 🟨 Word Embeddings ↳ For each word, look up corresponding word embedding vector from a table of 22 vectors, where 22 is the vocabulary size. ↳ In practice, the vocabulary size can be tens of thousands. The word embedding dimensions are in the thousands (e.g., 1024, 4096) [3] 🟩 Encoding ↳ Feed the sequence of word embeddings to an encoder to obtain a sequence of feature vectors, one per word. ↳ Here, the encoder is a simple one layer perceptron (linear layer + ReLU) ↳ In practice, the encoder is a transformer or one of its many variants. [4] 🟩 Mean Pooling ↳ Merge the sequence of feature vectors into a single vector using "mean pooling" which is to average across the columns. ↳ The result is a single vector. We often call it "text embeddings" or "sentence embeddings." ↳ Other pooling techniques are possible, such as CLS. But mean pooling is the most common. [5] 🟦 Indexing ↳ Reduce the dimensions of the text embedding vector by a projection matrix. The reduction rate is 50% (4->2). ↳ In practice, the values in this projection matrix is much more random. ↳ The purpose is similar to that of hashing, which is to obtain a short representation to allow faster comparison and retrieval. ↳ The resulting dimension-reduced index vector is saved in the vector storage. [6] Process "who are you" ↳ Repeat [2]-[5] [7] Process "who am I" ↳ Repeat [2]-[5] Now we have indexed our dataset in the vector database. [8] 🟥 Query: "am I you" ↳ Repeat [2]-[5] ↳ The result is a 2-d query vector. [9] 🟥 Dot Products ↳ Take dot product between the query vector and database vectors. They are all 2-d. ↳ The purpose is to use dot product to estimate similarity. ↳ By transposing the query vector, this step becomes a matrix multiplication. [10] 🟥 Nearest Neighbor ↳ Find the largest dot product by linear scan. ↳ The sentence with the highest dot product is "who am I" ↳ In practice, because scanning billions of vectors is slow, we use an Approximate Nearest Neighbor (ANN) algorithm like the Hierarchical Navigable Small Worlds (HNSW).show more