Innovative & Passionate
Who is behind the success of our initiative?
The TUM.ai Member Journey
All of our active members embark on an exciting journey - working on projects, driving AI innovation, and helping shape the TUM.ai community. At the heart of it all is our community, which empowers every member to grow, take initiative, and turn bold ideas into reality. No matter your ambition, TUM.ai is here to help you pursue it and make a meaningful impact - together.
Batch Introduction
Kick off your journey with the Batch Introduction! During the onboarding weekend, you'll get introduced to TUM.ai, meet current and new members, and take part in social events - setting the foundation for your journey. The experience continues with a getaway trip designed to deepen connections with both previous and new members in an unforgettable setting.
Research Track
Join a dedicated team working on an Impact Project aligned with your skills and interests, applying AI to address real-world challenges in a collaborative environment. These projects are designed to contribute to the broader research and technology community through tangible outcomes such as academic publications or open-source contributions. Along the way, you'll have the opportunity to engage deeply with our vibrant community, develop new skills, and take part in learning opportunities, trips, and special events—all while experiencing the hands-on, team-oriented culture that defines TUM.ai.
Initiative Track
Join one of our core departments and become a driving force behind everything that makes TUM.ai stand out. In this track, you'll dive into exciting projects, collaborate with motivated peers, and help shape the future of our community. Whether it's launching new ideas, strengthening our network, or making the day-to-day magic happen - you'll be at the heart of it all, growing your skills while making TUM.ai better for everyone. Alongside your work, you'll have the chance to engage deeply with our vibrant network, develop new skills, and take part in learning opportunities, trips, and special events.
Growth Opportunities
After your first semester, your journey at TUM.ai doesn't end with your initial project. You'll have the opportunity to further shape the initiative by founding a strategic task force, joining a department or task force that's new to you or continuing in your current department, potentially taking on a Team Lead role.
Research Exchange (REX) Program
After completing one semester, you become eligible for the REX Program-your gateway to conducting research at world-renowned institutions like MIT, Harvard, or Camebridge. Through our strong alumni network, we'll help you find the right research topic, navigate the application process, and support you every step of the way. It's a unique chance to take your AI journey global and contribute to cutting-edge research on an international stage.
Alumni Program
Having been with TUM.ai for two or more semesters, you can join the Alumni Program, opening up opportunities for continued networking and collaboration.
Our Core Departments
Discover the diverse teams that make TUM.ai thrive. Each department plays a crucial role in our mission to shape the future of AI.
Makeathon
Organizing Munich's largest AI Makeathon - a 48-hour virtual challenge where teams develop real-world business cases with AI. Features workshops, talks, and expert mentorship.
Venture
Bridging the gap between ideas and successful AI startups. We foster innovation and entrepreneurial spirit within TUM.ai.
Industry
Connecting TUM.ai members with industry partners for real-world AI projects during lecture-free periods.
Education
Creating and delivering AI educational content and events for all skill levels - from beginner to expert.
Software Dev
Building and maintaining TUM.ai's digital infrastructure, from in-house tools to cloud services.
Legal & Finance
Ensuring TUM.ai's compliance and financial sustainability while maintaining our non-profit status.
Community
Building and nurturing our community through events, buddy programs, and recruitment.
Marketing
Shaping TUM.ai's public image and promoting our vision through strategic communication.
Partners & Sponsors
Building strategic partnerships and securing sponsorships to enable ambitious events and projects.
Our Impact Projects
Explore our latest research papers and projects that showcase the innovative work of our members.
IBM SS24 Research
We address a fundamental limitation in language models: their struggle with numerical reasoning. While these models excel at text generation, they lack natural inductive bias for handling numbers, especially in scientific contexts. Our solution introduces two novel loss functions that improve numerical accuracy by incorporating regression-like objectives into the training process. When applied to a standard T5 model, we demonstrate significant improvements in numerical accuracy across mathematical tasks.
MIT SS24 Research
We bridge the gap between computational predictions and laboratory synthesis in materials science. While machine learning has improved the prediction of stable material structures, translating these predictions into successful laboratory syntheses remains challenging. Our work introduces the first Reaction Graph Network (RGN) to model precursor interactions and predict synthesis conditions for inorganic reactions, significantly accelerating the path from computational predictions to material synthesis.
MIT WS24/25 Research
We present Retro-Rank-In, a novel framework for inorganic material retrosynthesis that overcomes limitations in traditional machine learning approaches. Unlike previous methods that rely on known precursors, our approach embeds target and precursor materials in a shared latent space and learns a pairwise ranker on a bipartite graph. This enables unprecedented generalization capabilities, as demonstrated by correctly predicting precursor pairs never seen in training, setting new state-of-the-art results in material synthesis planning.