AI Landscape
AI landscape¶
This work is licensed under a Creative Commons Attribution 4.0 International License.
Table: Dominant LLM models currently in public use
| Name | Creator | Application | Access | Publications |
|---|---|---|---|---|
 LLaMa |
Meta | LLM | by request | facebookresearch/llama, (Touvron et al. ) |
| Segment-Anything | Meta | Computer Vision | free | facebookresearch/segment-anything, (Kirillov et al. ) |
| LaMDA | LLM | by request | (Thoppilan et al. ) | |
| BARD | Search, chat | free | ||
| Bing | Microsoft | Search, chat | free | |
| ChatGPT | OpenAI | Chat | free, subscription | |
| DALLĀ·E | OpenAI | Computer Vision, Chat | free | openai/DALL-E, (Ramesh et al.) |
| Megatron-Turing NLG | NVIDIA | LLM | by request | NVIDIA/Megatron-LM, (Shoeybi et al. ) |
Image Analysis¶
Stable Diffusion & Image Segmentation Models
Stable Diffusion
Stable Diffusion was released by the CompVis Lab group at Ludwig Maximilian University of Munich.
Stable Diffusion models are available via HuggingFace
Diffusion models have two modes, forward and reverse. Forward diffusion adds random noise until the image is lost. Reverse diffusion uses Markov Chains to recover data from a Gaussian distribution, thereby gradually removing noise.
Stable Dffusion relies upon Latent Diffusion Model (LDM) ()
Example Image Generation Models
DALLĀ·E uses GPT to create imagery from natural language descriptions
MidJourney uses a proprietary machine learning technology, believed to be stable diffusion, along with natural langauge descriptions in the same way as DALLĀ·E and Stable Diffusion models. MidJourney is only available via Discord, and requires a subscription for premier access after a 30-day free trial.
Segment-Anything (Meta), Kirillov et al. , is a recently released image and video segmentation technology that allows you to 'clip' a feature from an image with a single click.
Example Video Generation and Segmentation Models
Meta's Segment Anything can instantly identify objects in complex images and videos. Built on the SA-1B dataset, one of the largest image segmentation datasets ever publicly released, it saves technicians time and helps generate new training datasets for more refined computer vision model development.
Understanding Embeddings
What are Embeddings? - Vicki Boykis - download PDF
Embedded space for geospatial applications:
Visualizing how embeddings can organize satellite imagery. Millions of points covering the state of Alabama move between their geographic position and their location in the embedding space. pic.twitter.com/Z6FtoMQ84B
— Caleb Kruse (@clkruse) May 15, 2023
Embedded space for natural language:
Programmer Q/A¶
Phind.com - is a search engine optimized for developers and technical questions with simple explanations and relevant code snippets from the web, drawing from sources like StackOverFlow.