Research & Education

DIM is an open-source tool specifically designed for deep representation learning, focusing on the estimation of mutual information. Its core functionality revolves around maximizing the information shared between inputs and outputs, which is a crucial aspect of developing robust and efficient AI models. The tool is built to be compatible with PyTorch, a popular deep learning framework, making it accessible to researchers and developers already working within that ecosystem. It is associated with research presented at the International Conference on Learning Representations (ICLR) in 2019, indicating its foundation in academic rigor and advanced AI methodologies.

Paytm Labs is a dedicated research and development division that leverages artificial intelligence and machine learning technologies to innovate within the financial products sector. Its primary objective is to create and enhance solutions for both individual consumers and various types of merchants. Notable achievements include the development of the Paytm Canada application and significant contributions to the PayPay mobile payment service. The company is committed to building impactful, industry-changing products for the market.

GNN4NLP-Papers provides a comprehensive, curated list of research papers specifically focusing on the application of Graph Neural Networks (GNNs) to Natural Language Processing (NLP) tasks. The repository meticulously collects and organizes papers that have been published at prominent NLP and machine learning conferences. This resource is designed to serve as a valuable tool for both researchers and practitioners who are keen on keeping abreast of the most recent developments and breakthroughs in the intersection of GNNs and NLP.

LLM4TS is a comprehensive collection of resources specifically curated for the application of Large Language Models (LLMs) and Foundation Models (FMs) in the domain of time series data. This project serves as a valuable hub for researchers and practitioners, offering a compilation of relevant academic papers and practical code examples. Its primary objective is to enhance understanding and utilization of LLMs for various time series analysis tasks, thereby accelerating advancements in this specialized field.

LLMZoo is a project designed to support the advancement and assessment of large language models (LLMs). It offers a comprehensive suite of resources, including datasets, pre-trained models, and robust evaluation benchmarks. The platform is particularly useful for AI researchers and engineers involved in the development, fine-tuning, and performance assessment of LLMs. A key feature is its support for replicating multilingual instruction-following LLMs, aiming to democratize access to technologies similar to ChatGPT by providing necessary resources and tools.

Memery is an open-source tool designed for searching large image datasets. It leverages natural language processing and computer vision capabilities to enable users to find specific images by describing them in human language. This functionality is particularly useful for navigating and extracting information from extensive image libraries, making it a valuable resource for professionals working with large visual data collections.

OpenNMT-tf is a robust sequence learning toolkit leveraging the TensorFlow framework. It specializes in neural machine translation (NMT) and is also applicable to a broader range of sequence-to-sequence problems. The toolkit provides support for various neural network architectures and incorporates diverse training techniques, making it a versatile tool for developing and experimenting with sequence models. It caters to both researchers and practitioners working in the field of natural language processing.

Openfederatedlearning is an open-source framework designed for federated learning. It facilitates distributed machine learning approaches, allowing multiple parties to collaboratively train machine learning models without the need to centralize their raw data. This framework is specifically built to support privacy-preserving machine learning initiatives. It caters to the needs of machine learning researchers and data scientists who require secure and distributed model training capabilities. The project is available on GitHub.

CliniSpan Health is an innovative AI-powered platform focused on improving diversity within clinical research. It strategically utilizes social media channels to engage and reach underserved communities, providing cultural education and fostering participation. The platform's core mission is to bridge the gap between clinical researchers and diverse populations, thereby expanding access to medical studies for individuals who might otherwise be overlooked. By leveraging AI and social media, CliniSpan Health aims to make clinical trials more inclusive and representative.

Awesome-LM-SSP provides a curated collection of resources dedicated to the safety, security, and privacy aspects of large models (LMs). The repository specifically highlights multi-modal LMs, encompassing vision-language and diffusion models. It serves as a continuously updated resource, with manually collected materials, to assist researchers and practitioners in keeping abreast of the latest developments and best practices concerning the trustworthiness of large models.

bioemu is an open-source tool designed for the scalable emulation of protein equilibrium ensembles. It leverages generative deep learning techniques to approximate the equilibrium distribution of structures for a protein monomer, given its amino acid sequence. The tool provides researchers with inference code and pre-trained model weights, facilitating the study and understanding of protein dynamics and structural variations. Its primary function is to generate structural ensembles that represent the various conformations a protein can adopt under equilibrium conditions.

CapsGNN provides a PyTorch-based implementation of the Capsule Graph Neural Network architecture. This open-source project is specifically tailored for addressing graph classification problems. It serves as a valuable resource for researchers and developers who are actively engaged in the field of neural networks, particularly those focusing on graph-structured data and its classification challenges. The tool aims to facilitate the application and exploration of Capsule Graph Neural Networks in various research and development contexts.

ChangeDetectionRepository is an open-source project that provides a collection of Python implementations for various change detection techniques. It encompasses both traditional algorithms and modern deep learning approaches, including methods such as SFA, MAD, SiamCRNN, and DSFA. Beyond just the algorithms, the repository also includes multi-temporal datasets, making it a comprehensive resource for research and development in the fields of change detection and remote sensing image interpretation. This repository aims to facilitate the exploration and application of different change detection methodologies.

Conceptual Captions is a valuable dataset comprising image-URL and caption pairs. Its primary purpose is to facilitate the training and evaluation of machine learning models specifically designed for image captioning. By utilizing this dataset, developers and researchers can enhance the performance and accuracy of their image captioning systems. It serves as a foundational resource for advancing research in both computer vision and natural language processing, offering a standardized collection of data for model development and benchmarking.

ConvE is an open-source tool designed to provide resources for Convolutional 2D Knowledge Graph Embeddings. It serves as a platform for research and experimentation in the field of knowledge graph analysis. The tool is specifically built to facilitate the study of relationships and entities within complex datasets, making it suitable for researchers. Its primary applications include tasks such as link prediction and graph completion, offering a robust framework for advancing understanding in these areas.

company-research-agent is an AI-powered research assistant designed to perform in-depth due diligence on companies. It operates using a sophisticated multi-agent framework, leveraging the capabilities of LangGraph and Tavily for its core operations. For AI inference, it integrates advanced models such as Google's Gemini 2.5 Flash and OpenAI's GPT-5.1. This tool is built to support comprehensive company analysis and is capable of processing information in multiple languages, making it versatile for various research needs.

Deep-residual-networks is an open-source repository that hosts the original ResNet models, including popular architectures like ResNet-50, ResNet-101, and ResNet-152. These models are specifically designed and optimized for various image recognition tasks. The repository acts as a fundamental resource for researchers and developers engaged in the fields of deep learning and computer vision. It significantly aids in the implementation, experimentation, and study of deep residual learning techniques, making advanced computer vision accessible.

dreamsim is an open-source tool designed to explore and learn new dimensions of human visual similarity. It leverages synthetic data to achieve this, providing a platform for researchers to investigate perceptual alignment and its impact on vision representations. The tool is primarily aimed at facilitating advanced research in artificial intelligence and computer vision, offering a resource for professionals in these fields to deepen their understanding and develop more perceptually aligned AI models.

Shami Innovation and Technologies Pvt Ltd (SHAMIIT) is an Indian startup dedicated to fostering an ecosystem that spans training, recruitment, and emerging technologies. The company is actively engaged in AI research, with a specific focus on developing indigenous Large Language Models (LLMs). Beyond LLM development, SHAMIIT is also creating practical AI solutions tailored for the agriculture sector and advancing sustainable water treatment technologies. Their work aims to integrate cutting-edge AI into various critical domains.

School of Core AI provides comprehensive education in the rapidly evolving fields of AI, Machine Learning, Data Science, and Generative AI. The institution focuses on delivering industry-relevant skills through a practical, project-based learning methodology. Its curriculum encompasses a wide range of topics, including data analytics, machine learning, deep learning, and full-stack data science. Learners can also pursue certifications in popular AI frameworks such as PyTorch and TensorFlow, preparing them for real-world applications and career advancement in the AI sector.

Sourcely.com currently appears to be a domain that is for sale, with all pages displaying a message indicating "Buy this domain. We’re getting things ready Loading your experience… This won’t take long." Based on the scraped live website content, there is no active AI tool or service available under this domain. The previous description suggested it was an AI-powered tool for generating citations and bibliographies, but this functionality is not present on the current website. Therefore, no features, pricing, or specific use cases can be derived from the live site.

awesome-deep-learning-papers provides a meticulously curated collection of the most cited deep learning papers published between 2012 and 2016. This resource is designed to serve as a valuable reference for researchers and students, offering a structured way to explore and understand the foundational and influential works that shaped the early advancements in deep learning. The repository aims to simplify the process of identifying key research, enabling users to delve into the core concepts and methodologies that underpin modern AI.

Awesome-knowledge-distillation is a comprehensive, curated list designed for researchers and practitioners in the field of machine learning. It centralizes resources related to knowledge distillation, a technique used for model compression and transfer learning. The list includes links to relevant academic papers, articles, and other materials, making it a valuable hub for anyone looking to explore or implement knowledge distillation methods.

Awesome-Knowledge-Distillation-of-LLMs provides a comprehensive, curated list of academic papers specifically on the topic of knowledge distillation for large language models (LLMs). The resource organizes these papers into distinct categories, including methods for knowledge elicitation and various distillation algorithms. It also delves into the practical applications of distillation, such as skill distillation and vertical distillation of LLMs. This makes it an invaluable resource for researchers and practitioners working in the field of Natural Language Processing (NLP) who are interested in optimizing and understanding LLM performance through distillation techniques.