Pick the topics you want, in the order you want them. Save the plan and take it to a Program Advisor.
Develop AI models to predict and track the movement of CO₂ stored underground for carbon sequestration efforts.
Analyze datasets on oil and gas production and seismic activity to determine whether human activity is influencing earthquake patterns.
Apply machine learning techniques to forecast seismic activity and optimize geothermal energy production.
Analyze subsurface datasets and build machine learning models to identify potential oil and gas reserves and predict economic viability.
Explore how quantum computing could enhance deep learning, gaining insight into both classical and quantum model architectures.
Explore advanced quantum linear algebra algorithms, gaining insight into their theoretical foundations and practical applications.
Investigate quantum circuits as analogs of classical neural networks, exploring their potential advantages and implementation on real quantum hardware.
Explore how to process and learn from quantum data using quantum computing, preserving its rich structure.
Design and benchmark logical operations for error-corrected qubits using fault-tolerant techniques.
Assess quantum error-correcting codes under realistic noise to ensure reliable computations.
Implement quantum neural networks to solve machine learning and data modeling tasks.
Use variational quantum circuits to model quantum systems for chemical and materials simulations.
Investigate the impact of game-based stressors on cognitive function using data collected from adolescent players.
Investigate the role of baseline cognitive processes in predicting changes to stress and anxiety using longitudinal data.
Use data to identify individual characteristics that account for differential outcomes in response to performance-related feedback.
Use data to identify variables most highly predictive of educational achievement (i.e., GPA).
Build and benchmark a shifted-window transformer (SWIN ViT) model to solve the image classification task from Topic 1, with enhanced efficiency.
Develop ViT-based segmentation models to extract key biological markers and evaluate treatment response in clinical trial settings.
Implement vision transformer (ViT) models to automate the classification of medical images and streamline patient recruitment for clinical trials.
Prototype low-cost diagnostic tools and explore how AI can accelerate their design and global deployment.
Apply AI to analyze the microbiome and its influence on immunity, inflammation, and behavior.
Use AI to compare and model single-cell data across species to understand how new cell types emerge.
Explore how AI models can decode the genetic logic and cellular behavior that drive tissue regeneration.
Apply foundational optimization techniques to solve core problems in robotics, from perception to motion planning.
Design and implement depth sensing pipelines to extract 3D information from camera data, enabling perception and modeling in robotics.
Implement algorithms for Simultaneous Localization and Mapping (SLAM) to enable mobile robots to navigate and map unknown environments.
Develop algorithms to determine the position and orientation of objects in 3D space, a critical component for robotic manipulation.
Detect indoor furniture from image data to enable spatial awareness for robots and smart home devices.
Train models to detect human posture (standing, sitting, etc.) to improve real-time human-machine interaction.
Develop a GenAI system that animates human motion from a static image using image-to-video generation models.
Reconstruct 3D models from 2D camera images by learning depth inference techniques fundamental to robotic perception.
Design and implement AI-driven control systems for self-driving and robotic applications.
Explore how AI can optimize semiconductor design, reducing development cycles and improving energy efficiency.
Develop AI-powered vision systems for smart glasses, enabling real-time object detection and interaction in wearable technology.
Simulate AI-driven lending decisions to understand how financial institutions evaluate borrower risk and creditworthiness.
Build a sentiment analysis system using LLMs to measure economic mood from news, social media, and public discussions.
Simulate monetary policy decision-making by using LLMs to analyze economic indicators and determine interest-rate strategies.
Build a virtual marketplace where large language model agents act as consumers and generate demand behavior under different market conditions.
Navigate consumption-investment challenges in incomplete markets, focusing on optimal investment strategies and risk management.
Calibrate models for pricing complex financial instruments, crucial for assessing risks and setting premiums accurately.
Apply machine learning to predict asset prices and returns, a core component of financial analysis influencing profit and loss directly.
Build frameworks for making decisions with multiple conflicting objectives, applying quantitative tools to evaluate tradeoffs in business contexts.
Analyze resource allocation decisions through classic and modern optimization frameworks, with emphasis on fairness and efficiency in operations.
Explore how to use online learning algorithms to optimize pricing and product strategy in real time based on demand and inventory constraints.
Select their own dataset and create a customized dashboard to explore a meaningful social or personal theme.
Use the COMPAS dataset to uncover fairness issues in criminal justice risk scores using visualization techniques.
Visualize job-level AI usage using the Anthropic Economic Index to explore how AI is transforming work.
Explore real-world failures in voice assistants and build dashboards to highlight usability, bias, and design flaws.
Investigate methods for preserving privacy when using or deploying AI systems, with a focus on practical implementation.
Build LLM-powered tools to automate routine cybersecurity tasks such as vulnerability scanning or penetration testing.
Explore how to evaluate and improve the safety, interpretability, and goal alignment of AI systems.
Study how to bypass safety controls and explore vulnerabilities in AI models through adversarial prompting and jailbreaking.
Create models to measure the semantic similarity between sentences, enabling advanced text understanding applications.
Develop NER models that identify and classify entities within text using encoder-based models and modern NLP workflows.
Build sentiment analysis models using encoder-based language models to classify and evaluate text data effectively.
Integrate physical simulations with neural representations to model realistic motion and material behavior.
This track focuses on generating high-quality 3D content from single-view images using advanced AI techniques like DreamGaussian.
Create realistic human avatars using neural rendering techniques, enhancing the fidelity of virtual human representations.
Learn to reconstruct dynamic scenes or objects from 2D images or video frames using advanced neural rendering methods.
Discover how AI can interpret invisible signals using futuristic sensors.
Use AI to create original images—from realistic faces to wild artistic styles.
Teach AI to recognize and label objects in images—just like in self-driving cars or security systems.
Build 3D worlds from 2D images using AI tools used in AR, VR, and film.
