Friends,
What an amazing launch from summer into autumn! Thank you to the volunteers, faculty, lecturers, sponsors, staff, and attendees who made our first AI Summer School a success. We started off the week with charges from Dean Krish Roy for “VALIANT to be the place that sparks ‘radical collaboration’” while applying “this revolutionary technology in ways that are ethical, inclusive and beneficial for everyone” and from Provost C. Cybele Raver to address “the grand societal challenges of our time.”
I think we took a bit of a risk in going for a multi-disciplinary introduction to AI research, from core linear algebra and optimization to implementation and ethics. Yet, we received enthusiastic feedback (with will give us much needed data for continuous improvement for future events).
In terms of feedback, I’m thankful that folks were able to get into the QR codes for check-in/feedback. The discussions on data augmentation and ethics in AI resonated strongly, with comments such as, “The ethics in AI talk was very interesting overall,” and “Learning about augmentation was really interesting, especially considering how much is necessary before the data is no longer itself.” Other comments included, “The overview of transformers was extremely useful and interesting,” while another appreciated the practical insights, saying, “I really liked learning about how some of the features in Amplify work.” Additionally, the Bayesian statistics lecture and insights into neural networks left lasting impressions. Through the feedback, I see that the hands-on approach and the depth of topics covered were clearly appreciated, and I am relieved.
With the remarkable energy, curiosity, and insight displayed during the week together, I am eager to see transformative ideas take flight. Let’s change the world by understanding it!
– Bennett
Visionary Design
Jason Routhier(Director of Marketing & Communications at Vanderbilt University and Art Director of the VALIANT swirl) delved into his team’s creative process behind the logo. “We immediately went to the infinity symbol and the Mobius strip—those are the center points where everything multidisciplinary converges,” he explained. This infinity loop, according to Routhier, symbolizes the essence of collaboration, iteration, and ideation, capturing the ever-evolving nature of AI.
Reflecting on the complexities of transitioning the 2D design into a 3D format, Jason remarked, “There’s so much beauty in that abstract thinking, moving from one dimension to the next.” He envisioned the logo as more than a static image, imagining that it could animate with “a soft pulsing glow” when idle, and display vibrant changes when interacting with data or people.
As for the next steps, Jason emphasized the importance of figuring out how to “animate the planes” and incorporate images effectively. His enthusiasm for integrating these animations with AI was palpable, as he imagined a dynamic logo that responds to its environment. Routhier’s vision extends to the logo’s interaction with viewers, suggesting that its behavior could change based on the number of people around it. “If one person approaches, you see a single element, but if three people are watching, the logo becomes more animated, almost happier as more people engage with it.”
Our logos are not just designs—they are representations of how we, as a community, are embracing the future of technology and creativity. VALIANT is about pushing boundaries and blending disciplines. Our visual identity subtly reminds us of this core mission. Jason’s passion for bringing together art, technology, and human interaction is evident, as his work helps us refine VALIANT into a vibrant community of interdisciplinary collaboration.
As AI Summer School wraps up, I’m eager to see the VALIANT swirl making its way across campus on t-shirts, hats, and name plates. We hope that you find it refreshing to connect with the people who shape the literal vision that centers our community.
Alumni Lookout
This story from the field is brought to us by Dr. Zhoubing Xu (PhD’16, Vanderbilt Electrical Engineering), Scientific Director at Johnson & Johnson. His work focuses on high dimensional biological data (imaging & omics) to identify distinguishable phenotypes and differentiated mechanisms of action for efficient drug discovery.After getting my PhD in Electrical Engineering from the MASI lab at Vanderbilt University in 2016, I started my industrial career at Siemens Healthineers. There, I specialized in developing and translating AI/ML algorithms into commercial solutions, such as auto-contouring for radiation therapy planning and disease detection for computer-aided analysis. Recently, I joined Johnson & Johnson and shifted my focus to drug discovery to speed up the identification and development of new therapeutics through various modalities of biological data.
AI techniques, like ChatGPT, SAM (code) , and DINOv2 (code), have gained lots of popularity with their impressive capabilities in natural language processing and computer vision tasks. These advancements have captivated not only AI practitioners, but also the general public, who are fascinated by the “emergent” behavior of these models, where AI seems to autonomously tackle a wide range of tasks. While much of the buzz focuses on the sophisticated design of these models, it’s crucial to recognize the equally important role of data strategy. ChatGPT, for example, stands out from its competitors largely due to the introduction of Reinforcement Learning from Human Feedback (RLHF). This approach involves collecting a relatively small but specific dataset to align the AI with human preferences, significantly enhancing its performance. Similarly, SAM employs a human-in-the-loop strategy to efficiently gather high-quality data. By collecting 1 billion masks on 11 million images, SAM covers a wide range of common scenarios in computer vision segmentation, ensuring robust performance. DINOv2, on the other hand, curated a dataset of 142 million images from an initial pool of 1.2 billion raw images. This dedicated selection process prioritizes quality over quantity, proving that a more refined, diverse dataset can significantly boost performance compared to simply amassing vast amounts of data.
The AI community today has a spirit of future nostalgia. While it’s clear that advanced AI models are essential, AI practitioners cannot overlook the data quality, diversity, and relevance. Returning to first principles, data guides problem-solving. It is important to effectively curate (code), diversify (code), and generalize (code) the data, ensuring the AI innovation remains grounded in robust data-centric strategies. The VALIANT folks are sitting in a great position with their strengths: great platform, exceptional AI talents, and strong connections with collaborators who generate valuable data. By understanding and properly handling data, they can identify breakthrough points in technical challenges, achieving better performance and efficiency. Recognizing the importance of data will allow them to develop an intentional strategy to collect relevant and diverse data, enabling them to address more broad and generalized problems effectively, and eventually drive high impact advancements in the field.
“What I value most from my time at Vanderbilt is the understanding that everyone is the architect of their own business, from experimental design to career development. Adopting a mindset focused on robust data strategies forms the foundation for building skyscrapers in the booming AI era.”
VALIANT Ventures
- Prof. Tonia Rex is leading a new $1.4 million U.S. Department of Defense grant, “Diagnosis of Mild TBI Spectrum via Assessment Battery and Machine Learning” with Prof. Bennett Landman.
- Can Luo, Yinchen Liu, and Prof. Maizie Zhou introduced VolcanoSV, a hybrid structural variant detection pipeline, which surpasses existing tools in accuracy and robustness for identifying complex genomic alterations using single-molecule long-read sequencing data, making it a valuable asset for human genomics studies.
- Prof. Maizie Zhou led an international team to comprehensively benchmark computational methods for spatial transcriptomics, providing crucial insights to guide the selection of optimal tools for genomic research.
AI Summer School
From Music City in the Tennessee Valley, we are thrilled to announce our first international partnership with SCADS.ai in Music City Germany in the Saxon Valley (Leipzig University). We look forward to building on the long history of Vanderbilt-Leipzig collaborations with academic exchanges and partnerships in AI. Join as a AI Scholar or AI Fellow to find out more.Inventive AI
We are just two months into our AI Scholar VALIANT Innovation Incentive Award, and creativity is blooming. Keep up the great work! Be sure to remember to use the “Inventor Portal” (top row, center link) rather than the simplified Concept Submission forms when you are ready to discuss an invention with the CTTC..
Continuing Connections
We are excited to introduce the AI Alumni program, a new initiatve designed to foster lasting connections among Vanderbilts graduates. This program is open to alumni who graduated or work in related areas to VALIANT’s charge. The program provides a platform to stay updated on the latest advancements in AI research. Applications are now open. More details will unfold soon.Audience Choice
Congratulations to Prof. Jesse Spencer-Smith for being selected (in an extremely competitive vote) as the winner of the “Audience Choice Award” at AI Summer School for his outstanding talk, “Transformer Architecture from Zero to 60”!.
VALIANT Attempt
VALIANT Attempt, the student group for undergraduate and graduate affiliates, is holding an open interest meeting to discuss how to best serve the community.
When: Wednesday, 9/4 at 5pm
Where: AI Lounge (FGH/Jacobs adjacent to FGH 385)
Transforming Brain Tractography
Antoine Théberge from Sherbrooke University (Québec, Canada) will present a discussion of “Tractography Algorithms with Global Information via Transformers” on Thursday, September 5th, 2024, at 4pm. .
Alchemists’ Corner
- Benchmarking clustering, alignment, and integration methods for spatial transcriptomics
- Parallel signatures of cognitive maturation in primate antisaccade performance and prefrontal activity
- mTG-Gelatin phantoms as standardized testbeds for skin biomechanical measurements with Myoton
- SCCNAInfer: a robust and accurate tool to infer the absolute copy number on scDNA-seq data
valiant@vanderbilt.edu | https://vanderbilt.edu/valiant | @VandyValiant | LinkedIN
“Vanderbilt” and the Vanderbilt logo are registered trademarks and service marks of Vanderbilt University.