Research & Education

Vantage Learning offers K12 education technologies, specializing in Adaptive Learning Environments® and automated assessment solutions. Its flagship product, IntelliMetric®, is an intelligent scoring system that emulates human scorers for essays, leveraging Artificial Intelligence, Natural Language Understanding, and Computational Linguistics. This system can score essays in over 20 languages, provide holistic and domain-specific scores, and identify non-legitimate responses. Vantage Learning also provides MY Access!® for writing improvement, professional development, and custom solutions, aiming to enhance student achievement and support educators.

WebWalkerQALeaderboard is a Hugging Face Space that provides a comprehensive leaderboard for evaluating the performance of various AI models. It focuses on tasks such as deep search, web traversal, and Retrieval-Augmented Generation (RAG) systems. Researchers and developers can use this tool to compare how different models perform in specific scenarios, offering insights into their strengths and weaknesses. The platform is designed to help users understand the current state-of-the-art in AI model performance for complex question-answering tasks, making it a valuable resource for academic research and development in natural language processing and AI agents.

Whisper Cantonese Demo is an AI-powered tool designed for transcribing Cantonese speech to text. Leveraging the Whisper model, it provides robust speech recognition capabilities specifically for the Cantonese language. Users have the flexibility to input audio through various methods, including direct microphone input, uploading audio files, or even providing a YouTube video link. The tool processes the spoken Cantonese and outputs the transcribed text, making it highly useful for individuals who need to convert Cantonese audio into a written format. This includes language learners, researchers studying Cantonese, or anyone requiring accurate text transcriptions of Cantonese audio content.

Whisper Model Speech To Text is an AI-powered tool hosted on Hugging Face Spaces, designed to convert spoken language into written text. It leverages the advanced Whisper model to deliver accurate and efficient transcription services. Users can upload audio files to the platform and receive corresponding text outputs, making it suitable for a variety of applications requiring speech-to-text conversion. While the tool itself is a Hugging Face Space, the underlying infrastructure and advanced features are provided through Hugging Face's paid plans, offering options for increased storage, compute power, and dedicated inference endpoints. This makes it a versatile solution for individuals and teams looking for robust speech transcription capabilities.

ResearchPal is an AI-powered workspace designed to significantly accelerate academic writing and research tasks. It offers a suite of features including the generation of references, assistance with writing literature reviews, and tools to improve writing tone. Users can also read and interact with PDFs directly within the platform. ResearchPal supports integration with Zotero for seamless reference management and utilizes OpenAI Legacy Models, with access to more models available in its Standard and Pro subscriptions. The tool provides a free plan for users to test its features, ensuring accessibility for those looking to enhance their academic productivity.

Yuta by Ametros Learning is an AI-powered platform designed to revolutionize assessment in higher education. It helps universities and colleges design, deliver, and defend authentic competency-based assessments, addressing challenges like shrinking resources, inconsistent grading, and the rise of generative AI. Yuta automates assessment creation, delivery, and feedback, ensuring rigor and auditability. It features clear, reliable rubrics that can save up to 80% of grading time on major assignments, allowing instructors to focus more on teaching. Every assessment is mapped to course outcomes, providing personalized and actionable feedback to students. Built with a decade of experience in AI for education, Yuta prioritizes enterprise-level security, is SOC 2 compliant, and ensures data integrity by not feeding course content or student work into large language models.

Mondai | House of AI is the innovation center of TU Delft, dedicated to fostering collaboration in AI, Data, and Digitalization. It acts as a central hub, connecting research, education, businesses, and society to address complex technological and societal challenges through AI. The platform organizes events that bring together experts, talent, and professionals to share knowledge and initiate partnerships within the AI ecosystem. Mondai actively works on smart collaborations and AI partnerships to shape the future of AI in the Netherlands and Europe. It also highlights various AI projects with real-world impact, focusing on areas like Energy & Sustainability, Ethics & AI, FinTech, Generative AI, Health & Care, and more, demonstrating its commitment to responsible and impactful AI development.

The Robotics and AI Institute (RAI Institute) is dedicated to advancing the fields of robotics and artificial intelligence by tackling fundamental research challenges. Their mission is to create future generations of intelligent machines that are smarter, more agile, and easier to use, ultimately enhancing productivity, freeing humans from dangerous tasks, and assisting the disabled. The institute focuses on five core research areas: dexterous manipulation, advanced learning for control, data-driven AI models for physical robot interaction, navigation in challenging environments, and the ethics of robots in society. They publish research papers, blogs, and videos showcasing their work, including projects on dynamic whole-body manipulation and ultra-mobility in wheeled robotic systems.

OMNIOUS.AI is a research-focused company dedicated to exploring the vast potential of AI to augment human creativity and enhance personalization across diverse domains. Initially, the company leveraged AI to transform customer experiences within the fashion and e-commerce sectors. Now, OMNIOUS.AI is expanding into new areas with generative AI, unlocking creative possibilities for humanity. A key offering is Vella AI, which enables the creation of photorealistic virtual try-on images. Additionally, OMNICOMMERCE provides AI suites designed to deliver personalized e-commerce experiences. The company's small, dedicated team of developers and researchers is driven by the belief that AI can be a catalyst for human potential in ways yet to be imagined.

GPT-JT is an AI chatbot tool developed by togethercomputer, available as a Hugging Face Space. It allows users to generate text by providing prompts and selecting from a range of examples, including question answering and sentiment analysis. This tool is designed for testing and experimenting with language models, making it suitable for research, development, and general exploration of AI capabilities. While the live website currently shows a runtime error due to storage limits, its core functionality is to provide an interactive environment for users to engage with and understand the outputs of a language model.

MageBench Leaderboard is a platform developed by Microsoft, hosted on Hugging Face, designed to track and compare the performance of Large Language Models (LLMs). It functions as a public leaderboard where users can view benchmark results for various LLMs. Beyond just viewing, the tool also allows users to submit their own evaluations, providing details such as the score achieved, the name of the evaluation, the base model used, the testing environment, and the target research area. This makes it a valuable resource for researchers, developers, and anyone interested in understanding the current state and advancements in LLM performance. The platform is integrated within the Hugging Face ecosystem, leveraging its infrastructure for hosting and community interaction.

Moltbook is a unique social network designed specifically for AI agents, providing a platform where they can share information, engage in discussions, and upvote content. This environment allows for the observation of AI agent interactions, offering insights into their behavior and collaborative capabilities. The platform also features 'Submolts' and tracks live activity, including verified AI agents and posts. Humans can participate by sending their AI agents to Moltbook, following a verification process that includes signing up and tweeting to verify ownership. Additionally, Moltbook aims to support developers by allowing AI agents to authenticate with external applications using their Moltbook identity.

GovAI (Centre for the Governance of AI) is a research organization dedicated to understanding and shaping the future of advanced artificial intelligence. It provides rigorous research and fosters talent to help decision-makers navigate the complex landscape of AI development and deployment. The platform offers in-depth analysis on topics such as technical AI governance, AI progress forecasting, dual-use AI capabilities, and the societal implications of AI. GovAI publishes annual reports, research posts, and commentary, making its findings accessible to a broad audience. Its work has been featured in various publications, highlighting its influence in the field of AI governance.

Invisible Threads is an AI-powered platform designed to help users find the signal in the noise by discovering hidden connections between ideas across podcasts, essays, and books. It leverages AI-powered insight extraction to analyze content from sources like Lenny's Podcast and Paul Graham's Essays, identifying and visualizing the invisible threads linking related concepts. The tool provides thread visualizations to connect ideas, offers direct links to source content with timestamps for easy reference, and includes a "Rays" feature for random insight discovery. This makes it an invaluable resource for anyone looking to deepen their understanding and connect disparate pieces of information from diverse media.

DoLa (Decoding by Contrasting Layers) is an official implementation for a research paper focused on enhancing the factuality of large language models (LLMs). This open-source tool addresses the common problem of hallucinations in LLMs by introducing a novel decoding strategy. Instead of relying on retrieved external knowledge or additional fine-tuning, DoLa contrasts the differences in logits obtained from projecting later layers versus earlier layers to the vocabulary space. This method leverages the understanding that factual knowledge in LLMs is often localized to particular transformer layers. The approach has been shown to improve truthfulness across multiple-choice tasks and open-ended generation, with LLaMA family models seeing a 12-17% absolute point improvement on TruthfulQA.

Open-WebUI Tools is a modular toolkit designed to significantly enhance and enrich your Open WebUI instance, transforming it into a powerful AI workstation. This comprehensive collection includes over 20 specialized tools, function pipelines, and filters. It supports a wide range of applications, from academic research with arXiv search and Monte Carlo Tree Search, to multimodal creativity with advanced image and music generation tools like ComfyUI and Hugging Face Image Generator. The toolkit also facilitates agentic autonomy through tools like Planner Agent v3 and Letta Agent, alongside robust workflow capabilities. Key features include plug-and-play functionality, seamless visual integration with ComfyUI, intelligent model selection via Semantic Router, and full document processing capabilities.

AI Sweden is the national center for applied artificial intelligence in Sweden, dedicated to accelerating the adoption and application of AI for the benefit of Swedish society, competitiveness, and its citizens. As a non-profit organization, it is broadly funded and collaborates with over 170 partners from private and public sectors, as well as academia. The platform provides extensive resources including courses and competence development, AI labs for technology infrastructure and language technologies, and various sector-specific initiatives in areas like healthcare, energy, and automotive. It also fosters an ecosystem through networks, talent programs, and startup support, aiming to build AI capability and address critical challenges like cybersecurity.

Scira, formerly MiniPerplx, is a minimalistic yet powerful AI-powered search engine designed to streamline research. It leverages agentic planning to break down complex questions into sub-tasks, searches live sources, and cross-checks evidence to provide grounded answers with inline citations. Users can ask questions, upload PDFs, or paste URLs, and choose from 17 search modes including Web, Academic, Code, and X (Twitter) search. The platform also offers 28 tools for search and retrieval, financial data, location, media, and productivity. Scira supports a wide range of LLM models from providers like xAI, OpenAI, Anthropic, and Google, and is open-source under the AGPL-3.0 license, allowing for self-hosting and custom integrations.

Plan-and-Solve-Prompting offers a method for enhancing the zero-shot chain-of-thought reasoning capabilities of large language models. This open-source project, based on an ACL 2023 paper, provides the necessary code and resources for researchers and developers to implement and experiment with advanced prompting strategies. Notably, the Plan-and-Solve Prompting technique has been integrated into the core LangChain library as 'Plan-and-Execute,' highlighting its significance and utility in the field. Users can run the prompting methods with OpenAI API keys, either individually or with multiple keys for faster inference, and explore various prompt types like CoT, PS, and PS+ with different trigger sentences.

xLAM is a comprehensive platform offering a family of Large Action Models (LAMs) designed to enhance AI agent systems. It aggregates agent trajectories from diverse environments, standardizing them into a consistent format for optimized agent training. The platform includes various models, such as Llama-xLAM-2-70b-fc-r and xLAM-2-1b-fc-r, which are fine-tuned for broad agentic capabilities and specialized function calling tasks. xLAM also provides ActionStudio, a lightweight framework for agentic data and training, and APIGen-MT for multi-turn data generation. The models are compatible with VLLM, FastChat, and Transformers-based inference frameworks, making them suitable for researchers and developers looking to deploy and interact with advanced AI agents.

Transformer Lab is an open-source machine learning platform designed for AI researchers, unifying fragmented AI tooling into a single, elegant interface. It supports seamless training, evaluation, and scaling of models from local hardware to GPU clusters. Available in individual and team editions, it caters to researchers working on a single machine with local privacy and full toolkit access, as well as labs scaling across GPU clusters with features like unified orchestration for Slurm or SkyPilot, collaborative experiment tracking, and interactive compute sessions. Key capabilities include universal support for foundation models and LLMs, various training and fine-tuning methods, diffusion and image generation, and comprehensive evaluation and analytics.

London.AI is a meetup series presented by Air Street Capital and Nathan Benaich, focusing on both research and applied AI topics. These events gather members of the research, product, engineering, and founder communities to discuss advancements and applications in artificial intelligence. The meetups feature speakers from various AI-first companies and include networking opportunities. Air Street Capital, a venture capital firm investing in AI-first companies, also publishes the State of AI Report and Air Street Press, further contributing to the AI ecosystem. The events are designed to foster collaboration and knowledge sharing within the London AI community.

The MBZUAI Incubation and Entrepreneurship Center (IEC) is an AI-native incubator launched at the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI). It is dedicated to fostering the growth of AI startups and assisting innovators in bringing their AI-driven ideas to fruition. The center uniquely combines deep technical AI expertise with crucial strategic business insights, aiming to incubate value-creating startups. It provides a supportive ecosystem for entrepreneurs, leveraging the university's research capabilities and academic environment to accelerate the development and market entry of innovative AI solutions. The IEC is part of MBZUAI's broader 'Innovate' initiative, which also includes consulting and various institutes focused on specific AI applications.

Raycaster is an AI-native workspace specifically designed for the biopharmaceutical industry, aiming to accelerate drug development by streamlining document management. It ensures that protocols, reports, Module 2/3, and quality documents remain consistent and synchronized with the underlying scientific data. The platform offers features like regulatory research across various documents, AI-powered drafting of protocols and reports, and impact analysis to trace changes across the entire dossier. Raycaster helps teams spend less time on document reconciliation and more on scientific interpretation, ultimately leading to stronger submissions with significantly reduced rework for clinical, CMC, regulatory, quality, and nonclinical teams.