Research & Education

SeeMuseums is an innovative AI-powered mobile application designed to transform the museum visit experience. It functions as a personalized tour guide, leveraging artificial intelligence to create engaging and tailored journeys for each user. The app provides AI-driven guidance, helping visitors navigate museums, discover exhibits, and learn about art and artifacts in a more interactive and meaningful way. Its core purpose is to make museum exploration more accessible, informative, and enjoyable for a wide audience.

Evaluados AI is an innovative EdTech platform designed to support universities by generating high-quality academic content and resources using artificial intelligence. The platform focuses on enriching the educational experience through gamified learning and formative assessments, making learning more engaging and effective. Additionally, Evaluados AI provides valuable data-driven insights to institutions, enabling them to identify areas for educational improvement and offer personalized learning paths to students. This comprehensive approach aims to optimize both teaching methodologies and student outcomes within higher education.

Random Daily Art AI is a free daily newsletter that brings AI-generated art directly to your inbox. This service aims to provide a consistent source of artistic inspiration for its subscribers. By delivering new AI art pieces daily, it helps users explore and discover various AI art styles and techniques. It is particularly well-suited for art enthusiasts and creatives looking to broaden their artistic horizons and stay updated with the latest in AI-generated art.

Safe Sign Technologies is focused on creating a specialized Large Language Model (LLM) tailored for the legal domain. The primary goal is to address and solve the 'AI trust problem' within legal applications, thereby making reliable legal advice more accessible to a broader audience. The team behind Safe Sign Technologies comprises individuals with backgrounds from prestigious institutions such as Cambridge, MIT, Oxford, and Harvard, indicating a strong academic foundation for their work.

awesome-llm-skills offers a comprehensive, curated list of resources specifically designed for customizing AI agent workflows. This repository includes a variety of tools and skills that can be integrated with platforms such as Claude Code, Codex, and Gemini CLI. Its primary focus is on enabling developers and researchers to effectively customize and enhance their AI workflows. The resource is tailored for those actively working with Large Language Models (LLMs) and seeking to optimize their agent-based applications.

basic-memory is an open-source solution specifically designed to equip AI conversations with memory. This tool enables artificial intelligence systems to remember and reference previous interactions, fostering more coherent and contextually aware dialogues. It offers cross-device and multi-platform compatibility, ensuring flexibility in deployment. Additionally, basic-memory includes functionality for local knowledge graph storage, providing a robust mechanism for retaining and organizing conversational data.

bio_embeddings is a specialized tool designed to generate protein embeddings directly from protein sequences. This capability allows users to rapidly predict the structure and function of proteins based on their sequence data. The tool aims to simplify complex bioinformatics tasks by providing an efficient method for protein analysis. It offers comprehensive resources, including detailed documentation and a chat platform, to support users in learning and utilizing its features. bio_embeddings is also available as open-source code, promoting transparency and community contributions.

Bayesian-machine-learning is a repository of notebooks dedicated to exploring Bayesian methods within the field of machine learning. It serves as an educational resource, offering insights and practical examples to facilitate a deeper understanding of Bayesian machine learning concepts. The collection includes various examples and explanations of how Bayesian models work, making complex topics more accessible for learners. It is primarily designed for educational purposes, aiming to support individuals in grasping the fundamentals and applications of Bayesian techniques.

Awesome_GPT_Super_Prompting is a comprehensive collection of resources dedicated to the intricate world of GPT prompt engineering and its associated security implications. The repository offers valuable information on various aspects, including methods for ChatGPT jailbreaks, instances of prompt leaks, and detailed explanations of prompt injection techniques. This makes it an essential resource for both researchers and practitioners who are keenly interested in understanding the security vulnerabilities and robust engineering practices surrounding large language models and prompt design.

Story Crafter is a platform designed to enhance creative writing skills and support career development for writers. It features interactive lessons covering various genres, including fantasy, mystery, and poetry, allowing users to explore different writing styles. A key component is its AI-powered feedback system, which helps users refine their writing. Additionally, Story Crafter assists with the creation of professional documents like CVs, bridging the gap between creative pursuits and career opportunities.

Caffe is an open-source deep learning framework developed by Berkeley AI Research (BAIR). It emphasizes speed, expression, and modularity, making it suitable for a wide range of deep learning tasks. The framework is particularly well-suited for research and application development in areas such as computer vision. Caffe provides comprehensive tutorial documentation and DIY deep learning resources to help users get started and develop their projects effectively.

CADL is an open-source repository designed to support learning in "Creative Applications of Deep Learning with TensorFlow." It provides comprehensive course materials, including detailed lecture transcripts and practical homework assignments, all presented as interactive Jupyter Notebooks. This resource is ideal for individuals looking to learn and experiment with deep learning concepts using TensorFlow. The repository also includes a dedicated Python package containing code developed throughout the course, facilitating hands-on practice and deeper understanding of the subject matter.

captcha_recognize is a specialized tool focused on the image recognition of CAPTCHAs. Its core functionality involves recognizing CAPTCHAs without requiring prior image segmentation, leveraging the TensorFlow framework for its machine learning capabilities. The tool is implemented in Python 2.7 and is specifically designed to operate within an Ubuntu 16.04 environment. It is available as an open-source project, suggesting it can be freely used and modified by developers and researchers interested in CAPTCHA bypass or security research.

GenCode Labs offers AI solutions focused on SEO, machine learning, deep learning, and general AI technologies. The company assists businesses in utilizing data and artificial intelligence to improve their growth trajectories and operational efficiency. Their service offerings encompass custom AI solution development, strategic SEO planning, and the development and training of AI models tailored to specific business needs. GenCode Labs aims to be a partner for organizations looking to integrate advanced AI capabilities into their operations.

channel-pruning is an open-source initiative dedicated to enhancing the efficiency of deep neural networks. It achieves this by implementing channel pruning techniques, which are designed to significantly reduce the computational overhead associated with very deep neural networks. The project offers valuable resources, including those related to the ICCV'17 paper on channel pruning, making it a relevant tool for researchers and developers working on optimizing AI models.

brainstorm was an open-source neural network framework that offered tools for developing and experimenting with neural networks. It allowed users to build and test various neural network architectures. The project is no longer actively maintained, and its creators recommended transitioning to more current and supported frameworks such as TensorFlow or Chainer. These alternative frameworks provide enhanced features and improved performance, making them more suitable for contemporary AI development.

causal-ml provides a comprehensive, curated list of essential papers and resources for anyone interested in the intersection of causal inference and machine learning. This repository is designed to be a central hub for researchers and practitioners, offering valuable materials across various sub-topics. The collection includes resources on surveys, individual treatment effects, representation learning, and policy learning. Inspired by similar successful projects like GNNpapers, causal-ml also encourages community contributions to keep the list updated and expansive.

The research paper, 'CAMEL: Communicative Agents for “Mind” Exploration of Large Scale Language Model Society,' delves into the theoretical aspects of multi-agent interactions among Large Language Models (LLMs). It specifically investigates the dynamics and emergent behaviors of communicative AI agents when placed within simulated societal structures. This paper is intended as a foundational resource, providing critical insights for researchers and developers interested in the design and understanding of advanced AI agent systems and their complex interactions.

This foundational paper introduces the Generative Pre-trained Transformer (GPT) model, showcasing a new method for natural language understanding. It leverages unsupervised pre-training on a diverse text corpus to create powerful language models. These models can then be fine-tuned for various downstream Natural Language Processing (NLP) tasks, requiring only minimal task-specific data. This approach significantly advanced the field of language modeling and laid the groundwork for future developments in AI-driven language understanding.

ControlNet is a neural network architecture specifically designed to enhance the control over diffusion models, particularly in the context of text-to-image generation. It achieves this by introducing additional conditions, giving users more precise manipulation capabilities over the output images. This tool is highly beneficial for individuals working in AI art and research, providing a method to guide the creative process and experiment with controlled image synthesis.

AiSense is a platform that leverages artificial intelligence and machine learning to revolutionize disease diagnostics and healthcare. Its primary goal is to improve the precision of diagnostic processes, refine treatment plans, and expedite advancements in medical research. The platform includes a bio-signal monitoring system, facilitating patient registration, managing consent, and providing real-time health tracking capabilities. Additionally, AiSense offers training programs to support its users.

CS224n is a set of assignments designed for the Stanford CS224n course, focusing on Natural Language Processing with Deep Learning. These assignments delve into fundamental topics such as softmax, the basics of neural networks, word2vec implementations, and sentiment analysis. The exercises are structured to provide hands-on experience with core concepts in the field, requiring Python 2.7 and TensorFlow r1.2 for execution. It serves as a practical resource for students and learners interested in deep learning applications for NLP.

complexPyTorch is an open-source library designed to facilitate the use of complex-valued neural networks within the PyTorch framework. It offers a suite of tools and functionalities specifically for handling complex tensors and performing operations relevant to complex-valued neural networks. The library was particularly useful for developers working with complex numbers in neural networks prior to PyTorch's native support, streamlining the development process and enabling more advanced computational models.

course-content-dl is an open-source educational resource specifically designed for the Neuromatch Academy Deep Learning syllabus. It provides users with practical, code-first experience in understanding and applying deep learning theories and models. The resource emphasizes the critical thinking required to identify suitable problems for deep learning solutions and to select the most effective model implementations. Its primary goal is to support both practical applications and the advancement of scientific understanding in the field of deep learning.