Research & Education

MOSS-Speech Demo is an innovative speech-to-speech language model developed by the OpenMOSS-Team, available as a Hugging Face Space. This application enables users to input any text and receive an audio output spoken in a clear, human-like voice. The system generates an audio file that can be played directly or downloaded for later use. It is designed for experimenting with true speech-to-speech translation, making it suitable for research and development in multilingual communication. The tool provides a straightforward interface for quick text-to-speech conversion.

Intology is a research lab based in San Francisco dedicated to automating the process of scientific discovery. They develop end-to-end automated research systems, which they refer to as "Artificial Scientists." These systems have demonstrated significant capabilities, including publishing fully AI-generated A* conference papers and outperforming human experts in AI research and development tasks. Intology's core mission is to advance AI research through sophisticated automation, pushing the boundaries of what artificial intelligence can achieve in scientific exploration and discovery.

Netangular is an AI research and deployment company dedicated to advancing artificial intelligence technology. The company emphasizes safety and practical utility in its development process, aiming to create AI solutions that are both effective and responsible. Netangular values diverse perspectives and experiences, striving to make AI accessible and beneficial for addressing meaningful challenges across various domains. Their core mission revolves around building AI responsibly, ensuring that their innovations contribute positively to society while maintaining high standards of ethical development and deployment. The company's focus is on foundational AI research and its application to real-world problems.

Musicgen Songstarter Demo is an AI-powered tool hosted on Hugging Face Spaces, designed to help users quickly generate musical ideas. By providing a text description of the desired music, including genre, instruments, and tempo, the tool creates a 30-second stereo audio track. An optional feature allows users to upload a short melody, which the AI then uses as a guide to influence the generated output. This makes it an accessible platform for experimenting with different musical styles and overcoming creative blocks, providing a rapid prototyping solution for musicians and content creators.

Neuronika is a machine learning framework built entirely in Rust, emphasizing ease of use, rapid prototyping, and performance. At its core, Neuronika utilizes reverse-mode automatic differentiation, enabling the creation of dynamically changing neural networks with minimal effort and overhead through a lean, imperative, and define-by-run API. The framework leverages the power of the Rust language to offer an intuitive and efficient interface without the need for Foreign Function Interfaces (FFI). It supports GPU-accelerated primitives via CUDA, serialization with Serde, and transparent BLAS support for optimized matrix multiplication. Neuronika is currently in active development, with breaking changes expected as it evolves.

lagrangian_nns is an open-source project providing implementations of Lagrangian Neural Networks (LNNs). Unlike Hamiltonian Neural Networks, LNNs can parameterize arbitrary Lagrangians using neural networks and do not require canonical coordinates, making them suitable for systems where generalized momentum is difficult to compute, such as the double pendulum. The project includes a core equation of motion for automatic derivation of dynamics from learned Lagrangians, which is compatible with any JAX version. It also offers self-contained tutorials and example notebooks for various applications, including special relativity and the wave equation. The tool was developed with JAX 0.1.55 (2020) and uses pixi for reproducible environment management, ensuring compatibility with the original paper's environment.

KG-LLM-Papers is an open-source repository dedicated to curating a comprehensive list of research papers that explore the integration of knowledge graphs (KGs) and large language models (LLMs). This resource is invaluable for researchers and practitioners seeking to understand the synergies and applications at the intersection of these two advanced AI fields. The repository is actively maintained, with regular updates on new preprints and accepted papers from leading conferences. It categorizes papers into surveys and methods, offering a structured overview of the evolving landscape. The project encourages community contributions through Pull Requests to ensure the list remains current and complete.

NATSpeech is a comprehensive open-source framework for Non-Autoregressive Text-to-Speech (NAR-TTS) research and development. It offers official PyTorch implementations of advanced models like PortaSpeech (NeurIPS 2021) and DiffSpeech (AAAI 2022), facilitating high-quality and portable speech generation. The framework includes robust features such as data processing for NAR-TTS using Montreal Forced Aligner, a scalable training and inference system, and an efficient random-access dataset implementation. It's designed for technical users who want to explore and build upon state-of-the-art speech synthesis technologies, providing the necessary tools and code for experimentation and deployment.

LLM-Agent-Paper-List is a comprehensive repository of academic papers focusing on Large Language Model (LLM) based agents. This resource is specifically curated to accompany the 86-page SCIS cover paper, "The Rise and Potential of Large Language Model Based Agents: A Survey," by Zhiheng Xi et al. It offers researchers and developers a centralized and organized collection of must-read papers in this rapidly evolving field. The list is structured to cover various aspects of LLM-based agents, including their construction (brain, perception, action), practical applications (single-agent, multi-agent, human-agent cooperation), and the emerging concept of agent societies. The repository also includes news updates, project releases like AgentGym, and discussions on key topics and open problems, making it an invaluable tool for staying current with advancements in LLM agent research.

LM-reasoning is a GitHub repository dedicated to curating a collection of papers and resources focused on reasoning in Large Language Models (LLMs). It offers a structured overview of various techniques, including fully supervised finetuning, prompting and in-context learning, and hybrid methods, along with evaluation and analysis papers. The repository is designed to be a valuable resource for researchers, academics, and anyone interested in the advancements and methodologies behind reasoning capabilities in LLMs. It is open-source and encourages contributions from the community to ensure its comprehensiveness and up-to-date nature.

LLMAgentPapers is an open-source repository dedicated to curating a comprehensive list of must-read papers on Large Language Model (LLM) Agents. This resource is designed for researchers and practitioners in the field, offering an organized collection of academic works covering various aspects of LLM agents, including agent personality, memory, planning, tool use, RL training, and multi-agent systems. The repository also features sections on applications, frameworks, and benchmarks, making it a valuable hub for staying informed about the latest advancements and research trends in LLM agents. It provides direct links to paper abstracts, facilitating easy access to the research.

llm-continual-learning-survey is a comprehensive and actively updated survey focusing on Continual Learning of Large Language Models (CL-LLMs). This resource, published in ACM Computing Surveys 2025, serves as a dynamic repository of research papers, categorized by topics such as Continual Pre-Training (CPT), Domain-Adaptive Pre-Training (DAP) across various domains (legal, medical, financial, scientific, code, language), Continual Fine-Tuning (CFT), and more. It is designed to be a living document, with new papers and updates regularly added, and encourages contributions from the research community via pull requests or issues. This makes it an invaluable resource for researchers and academics tracking advancements in CL-LLMs.

Solomei AI is a research team dedicated to the exploration and integration of human creativity with artificial intelligence. The team comprises researchers from diverse fields including mathematics, engineering, philosophy, and the arts, fostering a multidisciplinary approach to AI development. Their core focus is on innovation that not only advances AI capabilities but also profoundly honors human values, ensuring that technological progress serves humanity's best interests. While specific tools or platforms are not detailed, the emphasis is on foundational research and ethical considerations within the AI landscape, suggesting a focus on theoretical and applied research rather than a direct end-user product.

Synboli revolutionizes material science by integrating cutting-edge AI with groundbreaking, patented chemistry to accelerate the discovery and development of high-performance polymers. The platform leverages proprietary synthesis technology to quickly produce and test polymer samples, enabling access to new, never-before-seen structures. Its AI-powered platform utilizes specialized models to rapidly design novel polymer structures and accurately predict their properties. By combining in-silico and in-laboratory evaluation, Synboli dramatically reduces polymer development time from years to mere months. This innovative approach allows for the creation of unprecedented materials with unique and tailored properties, addressing key needs in industries such as cosmetics, pharma, transportation, defense, construction, and electronics.

nlp_tasks is an open-source repository offering a curated collection of natural language processing tasks and selected references. It aims to provide a clear map of the NLP field, covering a wide array of tasks from Anaphora Resolution to Singing Voice Synthesis. The repository is continuously updated and encourages community collaboration through pull requests. It serves as an excellent starting point for researchers and practitioners looking to delve into specific NLP tasks, with references biased towards recent deep learning accomplishments. Each task entry includes relevant papers, projects, challenges, and datasets, making it a comprehensive resource for academic and practical exploration.

The Visual Computing Lab (VCL) at CERTH/ITI is a research entity focused on cutting-edge developments in visual computing. The lab specializes in creating advanced algorithms and architectures for various applications, including 3D and image/video processing. Its research scope extends to critical areas such as computer vision, pattern recognition, bioinformatics, and medical imaging. VCL actively contributes to the broader field of machine learning, pushing the boundaries of what's possible in intelligent systems and data analysis. The institute, founded in 1998, is a non-profit organization under the General Secretariat for Research and Technology, and a founding member of the National Centre for Research and Technological Development (EKETA).

This GitHub repository, Tanu-N-Prabhu/Python, serves as a comprehensive Open Source resource for learning Python and Machine Learning. It caters to individuals ranging from novices to seasoned developers, offering a structured path to mastery. The repository includes materials on basic Python concepts, built-in functions, popular libraries like NumPy and Pandas, and various APIs such as Google Translate and Wikipedia. It also delves into Machine Learning foundations, supervised and unsupervised learning, neural networks, and MLOps. Additionally, it provides extensive Data Science materials, including EDA techniques and real-world data analysis questions with Python answers. The resource emphasizes practical application through hands-on exercises and real-world examples, making it ideal for those looking to enhance their coding journey.

The python-machine-learning-book-2nd-edition repository serves as the official code and information resource for the second edition of the "Python Machine Learning" book. It provides comprehensive code examples, including Jupyter notebooks and Python scripts, for various machine learning algorithms and applications. Users can explore topics such as classification, dimensionality reduction, model evaluation, ensemble learning, sentiment analysis, regression, clustering, and deep learning with TensorFlow. The resource is ideal for students and professionals looking to implement machine learning concepts using Python, offering a practical, hands-on approach to learning.

python-ml-course is an open-source educational resource designed to introduce individuals to Machine Learning using Python. The comprehensive course covers a wide range of topics, from basic Python installation and data preprocessing to advanced concepts like Deep Learning and Reinforcement Learning. It includes practical exercises, real-world datasets, and all source code on GitHub, making it suitable for hands-on learning. The course is taught by Juan Gabriel Gomila, a professional in Data Science, and aims to make complex mathematical theories and algorithms accessible. It caters to students, programmers, and data analysts looking to specialize or enhance their skills in the lucrative field of Data Science.

RemoteCLIP is the official repository for the paper "RemoteCLIP: A Vision Language Foundation Model for Remote Sensing." This tool addresses limitations in existing remote sensing models by learning robust visual features with rich semantics and aligned text embeddings, crucial for retrieval and zero-shot applications. It leverages data scaling and conversion of heterogeneous annotations, incorporating UAV imagery to create a significantly larger pre-training dataset. RemoteCLIP supports diverse downstream tasks including zero-shot image classification, linear probing, k-NN classification, few-shot classification, image-text retrieval, and object counting, consistently outperforming baseline foundation models across various scales and datasets.

Physics-Informed-Neural-Networks (PINNs) is a research repository dedicated to investigating and implementing PINNs for solving Partial Differential Equations (PDEs). It integrates the physics of the PDE and boundary conditions directly into the neural network's loss function, utilizing the Mean-Squared Error of the PDE and boundary residual measured on 'collocation points'. The repository currently offers implementations for Burgers' and Helmholtz PDEs in both TensorFlow 2 and PyTorch. It also explores various aspects of PINNs, including the effectiveness of the L-BFGS optimizer for stiff PDEs, bottom-up learning mechanisms, and the impact of transfer learning on solution error, providing valuable insights for researchers and practitioners in scientific computing.

Practical Machine Learning with Python offers a structured and comprehensive three-tiered approach to learning machine learning and deep learning. This resource, based on a book, is packed with over 500 pages of useful information, helping readers master essential skills to recognize and solve complex problems with a data-driven mindset. It uses real-world case studies and leverages the popular Python Machine Learning ecosystem, including frameworks like scikit-learn, pandas, statsmodels, spaCy, nltk, gensim, tensorflow, and keras. The content covers machine learning concepts, the Python ecosystem, standard pipelines, and real-world case studies across diverse domains like retail, finance, and computer vision, making it ideal for practitioners.

PointLLM is a multi-modal large language model designed to understand colored point clouds of objects. It excels at perceiving object types, geometric structures, and appearance, effectively bypassing common issues like ambiguous depth, occlusion, or viewpoint dependency. The tool leverages a novel dataset comprising 660K simple and 70K complex point-text instruction pairs, enabling a robust two-stage training strategy. PointLLM also establishes two benchmarks, Generative 3D Object Classification and 3D Object Captioning, for rigorous evaluation. It offers capabilities for inferencing, chatting with 3D models, and evaluation using traditional metrics or GPT-4, making it a powerful resource for advanced 3D data analysis and robotics applications.

PointMamba is an open-source state space model (SSM) specifically designed for point cloud analysis, leveraging the success of Mamba from natural language processing. Unlike traditional Transformers, PointMamba employs a linear complexity algorithm, enabling global modeling while substantially reducing computational costs and GPU memory usage. This tool utilizes space-filling curves for efficient point tokenization and features a simple, non-hierarchical Mamba encoder as its backbone. Comprehensive evaluations demonstrate its superior performance across various datasets, making it a valuable resource for researchers and developers in 3D vision. PointMamba underscores the potential of SSMs in 3D vision-related tasks and provides a robust baseline for future research.