1/ 🔥 New paper: Differentiable Vector Quantization (DiVeQ) 🔥... Vector quantization (VQ) is a key building block in modern AI. It links continuous data like images and audio to discrete representations (tokens) used by transformers.show more

`transformers` + `torchao` quantization + `torch.compile` for faster inference... speed and less memory usage 🔥 Demo of "meta-llama/Meta-Llama-3.1-8B-Instruct" quantized in 4-bit weight-only :show more

A comprehensive map of the modern AI ecosystem, categorizing... key tools and platforms across models, agent frameworks, embeddings, security, observability, and vector databases. Features major industry names like OpenAI, Anthropic, Google, Meta, and LangChain. #AI #GenerativeAI #AIEcosystem #TechStack #LLM #MachineLearning #DataScience #TechTools #SoftwareDevelopment #ArtificialIntelligenceshow more

Fun fact: We actually built our AI emoji picker... Dub.co with Jev at first, but since we had to feed the entire emoji catalog into Jev, it consumed a lot of input tokens. Refactored to a simple vector index on Upstash vector, and costs became nearly negligible 🤯 Just shipped this to prod – h/t Pedro Ladeira 👏show more

Stokes' Theorem is a classic result from vector calculus.... It tells us that the line integral of a vector field over a loop is equal to the surface integral of the curl of the vector field over some enclosed surface.show more

This is how you can make a shape vector... illustration effect using figma & heatmap effect in Paper. Tutorial.show 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

MLX MiniMax 2.5 running LOCALLY on a single M3... Ultra 512GB! Writing a poem on LLMs at 6bit quantization! 🔥 Let's start some coding, context and distributed tests! Generation: 40.2 tokens-per-sec Peak memory: 186 GBshow more

1. GPT Images 2.0 でイラスト生成 2. Recraft V4 Vector... でSVG化 3. Adobe AI ターンテーブルで調整show more

Offset vector paths now available in Figma Draw. You... can now take any flattened vector path and offset it to expand or retract it. It even lets you round out those sharp edges.show more

A comprehensive visual map of the modern AI coding... agent ecosystem, highlighting 40 key tools spanning LLMs, coding assistants, agent frameworks, vector databases, infrastructure, and observability platforms. #AICoding #LLM #DevTools #CodingAgents #TechStack #SoftwareEngineering #MCP #VectorDatabase #Observability #ProductionReady #TechMapshow more

K-Means is simple. Making it fast on GPU isn't.... Flash-KMeans is an IO-aware implementation of exact k-means that rethinks the algorithm around modern GPU bottlenecks. By attacking the memory bottlenecks directly, Flash-KMeans achieves: - 30x speedup over cuML - 200x speedup over FAISS Using the same exact algorithm, just engineered for today’s hardware. At the million-scale, Flash-KMeans can complete a k-means iteration in milliseconds. Here's why this matters today: K-means has always been an offline primitive. Something you run once to preprocess data and move on. These speedups change that. ↳ Vector databases like FAISS use k-means to build search indices. Faster k-means means you can re-index dynamically as data changes, not batch it overnight. ↳ LLM quantization methods need k-means to find optimal weight codebooks, per layer, repeatedly. What takes hours could now take minutes. ↳ MoE models need fast token routing at inference time. Millisecond k-means makes it viable to run this inside the inference loop, not just in preprocessing. The 200x over FAISS is the number to internalize. FAISS is the industry standard. Most production vector search systems sit on top of it. Link to the paper and code in next tweet!show more

Excited to introduce #Paper2Agent: automatically transform passive paper📜 into... active AI agent🤖! Research paper ➡️ paper MCP ➡️ chat agent applying tools/data in the paper to new projects. Key is to use multiple sub-agents to build robust MCP from paper/code. Examples in🧵show more

K-Means is simple. Making it fast on GPU isn't.... Flash-KMeans is an IO-aware implementation of exact k-means that rethinks the algorithm around modern GPU bottlenecks. By attacking the memory bottlenecks directly, Flash-KMeans achieves: - 30x speedup over cuML - 200x speedup over FAISS Using the same exact algorithm, just engineered for today’s hardware. At the million-scale, Flash-KMeans can complete a k-means iteration in milliseconds. Here's why this matters today: K-means has always been an offline primitive. Something you run once to preprocess data and move on. These speedups change that. ↳ Vector databases like FAISS use k-means to build search indices. Faster k-means means you can re-index dynamically as data changes, not batch it overnight. ↳ LLM quantization methods need k-means to find optimal weight codebooks, per layer, repeatedly. What takes hours could now take minutes. ↳ MoE models need fast token routing at inference time. Millisecond k-means makes it viable to run this inside the inference loop, not just in preprocessing. The 200x over FAISS is the number to internalize. FAISS is the industry standard. Most production vector search systems sit on top of it. Link to the paper and code in next tweet!show more

Drag-and-drop UI to build AI agents! Langflow is a... powerful visual tool for building and deploying AI-powered agents and workflows—without writing any code. Supports all major LLMs, vector DBs, etc. 100% open-source, 62k+ stars 🌟show more

This is how you can create a super viral... 3D vector effect which makes few people question their existence. Built in Paper. Tutorial.show more

Vector is trying to be a good boy. #3DprintingFursuit

Diffusion Transformers aren't just generative models, but also powerful... multi-modal encoders. ConceptAttention creates rich heatmaps of text concepts in images from DiT representations. This even works on real images, and can be applied to tasks like segmentation! Demo 👇show more