Popular GitHub Repositories for Visualized Activations

This project extends the idea of the innovative architecture of Kolmogorov-Arnold Networks (KAN) to the Convolutional Layers, changing the classic linear transformation of the convolution to learnable non linear activations in each pixel.

Compare SELUs (scaled exponential linear units) with other activations on MNIST, CIFAR10, etc.

AG2 Visualized - Build Multi-Agentic Apps with Drag-and-Drop.

This is an open-source version of the representation engineering framework for stopping harmful outputs or hallucinations on the level of activations. 100% free, self-hosted and open-source.

RNN and general weights, gradients, & activations visualization in Keras & TensorFlow

Code accompanying the paper "Massive Activations in Large Language Models"

Quantize weights and activations in Recurrent Neural Networks.

Twitch Streamer GPT is a NodeJS-based Twitch enhancement tool, offering interactive stream experiences with AI-powered automated responses, voice command activations, and advanced modules. It's easy to set up and suited for users of all tech levels.

Compression schema for gradients of activations in backward pass

Evaluation code and datasets for the ACL 2024 paper, VISTA: Visualized Text Embedding for Universal Multi-Modal Retrieval. The original code and model can be accessed at FlagEmbedding.

Visualising what each LSTM cell learns from data.

A tool to visualize convolutional layer activations on an input image.

This is a master project of some experiments with Neural Networks. Every project here is runnable, visualized and explained clearly.

I developed 2 machine learning software that predict and classify ozone day and non-ozone day. The working principle of the two is similar but there are differences. I got the dataset from ics.icu. Each software has a different mathematical model, Gaussian RBF and Linear Kernel, and classifications are visualized in different ways. I would be happy to present the software to you!

A visualized text mining and analysis tool for student markdown reflection documents based on Natural language processing in the Dept of CS at Allegheny College.

Several datasets are manipulated, visualized, and analyzed with well-known ML Algorithms to make predictions, clustering, or classifications.

Quantization is a technique to reduce the computational and memory costs of running inference by representing the weights and activations with low-precision data types like 8-bit integer (int8) instead of the usual 32-bit floating point (float32).

This CNN-based language model leverages causal and dilated convolutions, gated activations, residual connections, and layer normalization to effectively model textual data for generation tasks.

This repository features an ETL pipeline for retrieving, processing, validating, and ingesting game metadata and sales data from SteamSpy and Steam APIs. Data is stored in a MySQL database on Aiven Cloud and visualized using Tableau dashboards for insightful analysis of gaming trends and sales performance.

A plug-and-play adapter architecture that efficiently adapts large language models to downstream tasks. Essential for client-server architecture language model services. A practical approach to adapting LLMs' hidden activations without modifying the base model.