Research & Education

The Centre for Artificial Intelligence and Digital Ethics (CAIDE) is a prominent research and teaching center based at the University of Melbourne. CAIDE is dedicated to advancing the understanding and responsible development of AI and digital ethics. Through its cross-disciplinary approach, the center aims to demystify complex topics in AI, fostering informed discussion and innovation. CAIDE actively supports researchers, academics, and policymakers by providing a platform for collaborative research, educational initiatives, and policy engagement. Its work contributes significantly to shaping the ethical landscape of artificial intelligence and its societal impact.

Artificial Intelligence Research and Development Lab (AIRnD Lab) is dedicated to empowering individuals and organizations with the skills and insights needed to thrive in the evolving world of AI. Incorporated in 2024, the lab focuses on fostering innovation and excellence in AI through a team of experienced researchers and educators. Its mission is to democratize AI education and research, making it accessible to everyone from beginners to advanced practitioners. AIRnD Lab organizes online and offline projects, research, and training programs for academics and working professionals, accelerating both research and skill development. The platform emphasizes bridging the gap between theoretical knowledge and practical application, enabling students and partners to harness AI for real-world problem-solving.

awesome-multi-agent-papers is a comprehensive compilation of leading research papers focused on multi-agent systems, curated by the Swarms Team. This resource aims to democratize access to key research in the field, making it easier for researchers, developers, and academics to stay updated on the latest advancements. The collection covers a wide range of topics including multi-agent collaboration and system design, frameworks and benchmarks, application-specific multi-agent systems in software engineering, healthcare, data & ML, multimodal applications, and other domains like urban planning and legal reasoning. It also includes papers on evaluation and model improvement for multi-agent LLMs, offering a valuable resource for anyone looking to explore or contribute to the evolving landscape of multi-agent AI.

awesome-multi-task-learning is a meticulously curated list designed for researchers and practitioners in the field of Multi-Task Learning (MTL). It compiles essential resources including datasets, codebases, and academic papers, all from a machine learning perspective. The project is continuously updated and greatly appreciates community contributions, ensuring its relevance and comprehensiveness. It covers various aspects of MTL, from benchmark datasets in computer vision and natural language processing to different architectural approaches like hard and soft parameter sharing, and optimization strategies. This resource is invaluable for anyone looking to explore, implement, or stay current with advancements in multi-task learning.

awesome-speech-recognition-speech-synthesis-papers is an open-source GitHub repository that serves as a curated list of academic papers focused on various aspects of speech technology. It covers key areas such as Automatic Speech Recognition (ASR), Speaker Verification, Speech Synthesis (TTS), Language Modelling, Singing Voice Synthesis (SVS), and Voice Conversion (VC). The repository is organized by topic, making it easy for researchers, academics, and students to find relevant literature. It includes papers ranging from foundational works to recent advancements, often providing direct links to PDF versions. This resource is invaluable for anyone looking to delve into the theoretical and practical developments in speech processing.

Awesome-Robotics-3D is a comprehensive, curated list of 3D Vision papers specifically focusing on the intersection of robotics and large models such as Large Language Models (LLMs) and Vision-Language Models (VLMs). Inspired by awesome-computer-vision, this repository serves as a valuable resource for researchers and academics. It categorizes papers into key areas like Policy Learning, Pretraining, VLM and LLM applications, Representations, and Simulations, Datasets, and Benchmarks. Each entry typically includes links to the paper, associated webpages, and code, making it easy for users to access and explore the research. The list is actively maintained and encourages contributions from the community.

awesome-nlp is a comprehensive, curated list of resources for Natural Language Processing (NLP) enthusiasts and researchers. It provides an extensive collection of links covering research summaries and trends, prominent NLP research labs, and various tutorials. The resource also includes reading content, videos, online courses, and books related to NLP. Furthermore, it categorizes libraries by programming language, such as Python, Node.js, C++, Java, and more, making it easy to find relevant tools. The list also highlights NLP resources specific to different languages like Korean, Arabic, Chinese, and German, catering to a diverse audience interested in the field.

awesome-nlp-sentiment-analysis is a comprehensive, open-source collection of resources for Natural Language Processing (NLP), with a strong emphasis on sentiment analysis. This GitHub repository compiles relevant datasets, academic papers, and practical open-source implementations. It specifically targets sub-areas such as general sentiment analysis, emotion cause recognition, and the extraction of evaluation objects and associated words. Researchers, students, and developers in the NLP field can leverage this curated list to find foundational knowledge, state-of-the-art research, and practical code examples to advance their work in understanding and processing human emotions and opinions from text.

Awesome-Jailbreak-on-LLMs is a comprehensive collection of resources dedicated to state-of-the-art jailbreak methods for Large Language Models (LLMs). This repository serves as a valuable hub for researchers and academics, offering a curated selection of papers, associated code, relevant datasets, and detailed evaluations. It covers various attack vectors, including black-box, white-box, multi-turn, and multi-modal attacks, as well as defenses and analytical frameworks. The project aims to foster understanding of LLM vulnerabilities, aiding in the development of more robust and secure AI systems. Contributions are welcome, making it a dynamic and evolving resource for the AI security community.

Humane AI Net is a prominent European research network dedicated to advancing human-centered artificial intelligence. The consortium actively funds a large community of AI scientists through smart funds, supporting both micro and macro-projects across various work packages including Human-in-the-Loop AI, Multimodality, Collaboration and Interaction, Societal AI, and Ethics and Responsible AI. It organizes numerous events such as conferences, workshops, and hackathons, fostering a vibrant community for interdisciplinary AI research. Humane AI Net also publishes reports and research, contributing significantly to the discourse on legal protection by design in the AI value chain and the role of AI metrics.

Awesome-Korean-NLP is a comprehensive, curated list of resources dedicated to Natural Language Processing (NLP) for the Korean language. This GitHub repository serves as a central hub for various tools, datasets, blogs, research papers, lectures, and online communities relevant to Korean NLP. It includes specific sections for morpheme/PoS taggers, named entity taggers, spell checkers, syntax parsers, sentimental analysis tools, translators, and general NLP packages. The resource also lists significant Korean datasets like Sejong Corpus and Wikipedia Dump, alongside academic papers and lectures from prominent institutions. It's an invaluable resource for anyone working with Korean language data, from academic researchers to developers building Korean NLP applications.

Awesome-Language-Model-on-Graphs is a comprehensive, curated list of papers and resources focusing on the intersection of large language models (LLMs) and graph structures. This repository is built upon the foundational survey paper, 'Large Language Models on Graphs: A Comprehensive Survey,' and serves as an evolving resource for researchers and academics. It categorizes research into areas such as pure graphs, text-attributed graphs, and text-paired graphs, detailing various approaches like LLM as Predictor, Graph As Sequence, and LLM as Aligner. The resource also highlights key datasets, direct answering methods, and heuristic and algorithmic reasoning techniques. It is designed to be continuously updated, making it an invaluable tool for staying current with advancements in this rapidly developing field.

awesome-online-machine-learning is a comprehensive, open-source curated list of resources dedicated to online machine learning. This field focuses on machine learning where data arrives sequentially, allowing models to update incrementally with one data point at a time, contrasting with traditional batch learning. The repository provides valuable links to courses, books, blog posts, and software related to online ML. It also features an extensive collection of research papers covering various online learning topics such as linear models, support vector machines, neural networks, decision trees, unsupervised learning, time series analysis, drift detection, and anomaly detection. This resource is ideal for anyone looking to deepen their understanding or find tools for online machine learning.

Monash DeepNeuron is a student team dedicated to the exploration and advancement of Artificial Intelligence (AI) and Optimised Computing (OC). Based at Monash University, the team focuses on empowering, engaging, and educating individuals in these fields. Through hands-on research projects, they equip students and researchers with essential skills and knowledge. The initiative fosters collaborative networks between students, academic staff, and industry partners to drive innovation and excellence. Monash DeepNeuron also emphasizes accessibility through community outreach events and educational initiatives, aiming to demystify AI and OC and promote understanding for broader participation in the digital revolution. Their diverse team includes members from various disciplines, working on projects across aerospace, science, medicine, gaming, and psychology.

The Learning Systems and Robotics Lab (LearnSysLab), formerly the Dynamic Systems Lab, is a research group led by Prof. Angela Schoellig at the Technical University of Munich and the University of Toronto Institute for Aerospace Studies. Their research is driven by the vision of seamless interaction between robotic systems and the physical world, particularly addressing challenges in unstructured, uncertain, and changing environments. They combine ideas from controls, machine learning, and optimization to develop next-generation robot algorithms that integrate a-priori information with operational data. Their work includes developing advanced robot control and learning algorithms for real-world applications and collaborating on interdisciplinary robot projects.

The EduAIhub is an African-based research network dedicated to fostering responsible artificial intelligence innovations to advance education across the continent. Located within the University of Lagos, the hub collaborates with institutions like Université d’Abomey Calavi and Data Science Nigeria. It brings together African researchers and innovators in AI and education, leveraging its network to forge partnerships with industry and academia. The hub supports high-quality research projects in three key thematic areas: inclusion, language, and administration, aiming to address educational challenges such as exclusion, disparities, and language barriers in Sub-Saharan Africa. The initiative also offers grants for projects developing AI solutions in these areas.

OdysseyGPT is an enterprise document intelligence and IDP platform designed to turn unstructured documents into reliable, citation-backed knowledge and structured data. It leverages retrieval-augmented generation, semantic search, and multi-step reasoning to extract data from various document types like contracts, invoices, resumes, and emails. The platform ensures data quality, transparency, and control by linking every extracted data point back to its exact source within the document. OdysseyGPT allows users to set up workspaces, roles, and approval steps, logging every action for auditability. It integrates with existing business systems like accounting, HR, CRM, and support tools, enabling seamless data flow while preserving source context. The tool is built with robust security features, including SSO, role-based access control, end-to-end encryption, and full audit trails.

Triomics offers a generative AI-powered platform specifically designed for oncology workflows, trusted by top academic and community cancer care providers. The platform, powered by its foundational oncology-tuned AI called OncoLLM, ingests various medical records (notes, PDFs, scanned faxes, labs) from EHRs. It then intelligently transforms complex, unstructured data into clean, real-time insights and actions. Triomics helps improve trial accruals by automatically screening appointments against trial portfolios, assists providers with pre-charting by assembling expert-level patient snapshots, and provides expert intelligence for data curation with high accuracy. This significantly reduces the manual effort required for reviewing medical records in oncology.

Irregular is a frontier AI security lab dedicated to protecting the world from the risks posed by advanced AI systems. The lab conducts research and evaluations, partnering with leading AI organizations, industry, and governments to address critical security challenges. Their work includes assessing AI models like GPT-5.5 and Meta’s Muse Spark against offensive security benchmarks, publishing findings, and contributing to the broader understanding of AI safety. Irregular is actively building a team of AI researchers, cyber researchers, and engineers to further its mission in AI security.

Awesome-MLLM-Hallucination is a comprehensive GitHub repository offering a curated list of resources focused on the phenomenon of hallucination in Multimodal Large Language Models (MLLM), also known as Large Vision-Language Models (LVLM). This resource is invaluable for researchers and practitioners, providing access to papers, code, and other materials. It categorizes resources into hallucination evaluation & analysis and hallucination mitigation, with papers listed from newest to oldest. The repository highlights key observations from recent updates, including diverse analyses of root causes, the rise of training-free mitigation solutions, the importance of contrastive decoding, and emerging RL-based methods. It also covers fresh angles for mitigation such as visual prompting, RAG, and rationale reasoning. The project is actively maintained and welcomes contributions.

Barie AI is an advanced AI agent designed to streamline complex workflows, from deep research and market analysis to strategy execution. It boasts the world's largest context window, enabling it to handle research tasks that other AI tools cannot, surfacing every source live. Barie AI integrates with numerous applications, allowing for multi-app workflows and automating tasks with contextual awareness. It also features a complete creation suite for generating videos, designing images, and building slides from a single prompt. Proven by GAIA Benchmark Testing, Barie AI adapts to various professions, offering capabilities like coding assistance, web search, and connectors to automate tasks, making work smarter and faster.

SynthLabs is an AI post-training research lab dedicated to advancing frontier AI capabilities. Their work centers on Reinforcement Learning from Human Feedback (RLHF), Reinforcement Learning from AI Feedback (RLAIF), and synthetic data generation. The lab adopts an open science approach to tackle post-training challenges, reinforcement learning, and AI alignment. Key research areas include scaling synthetic reasoning, developing generative reward models for better interpretability, and implementing Direct Principle Feedback to control language model behaviors without extensive retraining. SynthLabs aims to build more capable and trustworthy AI systems by exploring methods beyond human data for alignment at scale and preserving rich human intent.

Physics-Based-Deep-Learning is an open-source GitHub repository that compiles an extensive list of research papers and projects focused on the intersection of physical modeling and deep learning techniques. This resource is particularly valuable for those interested in "Physics-Based Deep Learning" (PBDL), covering methods that combine physical simulations with artificial neural networks. The repository categorizes approaches based on whether they target forward simulations or inverse problems, and by the integration type between learning and physics, including data-driven, loss-term, and interleaved methods. It emphasizes fluid flow and Navier-Stokes related problems, primarily featuring works from the I15 lab at TUM, alongside contributions from other research groups. This collection is ideal for academic research and staying updated on advancements in differentiable physics and AI for scientific computing.

The wuying-agentbay-sdk provides a cloud sandbox environment specifically designed for AI agents, enabling them to operate in isolated, on-demand settings. This SDK supports multiple programming languages including Python, TypeScript, Golang, and Java, offering a comprehensive API for agents to interact with a full cloud environment. Key functionalities include automating web operations, controlling cloud desktop applications, managing mobile UI automation, and providing a professional cloud development environment for code generation, compilation, and debugging. It eliminates the need for users to manage infrastructure, allowing agents to perform tasks and then be torn down, making it ideal for testing, development, and automated workflows.