CESA logoCESA

Projects

Members can join existing work or propose their own response to an environmental or social challenge.

Pollution over a city skyline

Sustainability

Modeling Pollution in the Seattle-Puget Sound Area

Model pollution exposure across the Seattle-Puget Sound region to better understand local environmental risks.

Ethical software audit

Tackling Unethical and Unsafe Software

Performing Automated Audits of Software to Create More "Ethical" Code

Develop repeatable audits that identify harmful patterns and encourage safer, more responsible software.

Computer science course planning

Increasing Digital Awareness

Creating Equitable Computer Science Syllabi for K-12 Students

Create accessible computer science learning materials that serve K-12 students from a wider range of backgrounds.

Machine learning weather forecaster

Sustainability

Creating a State-of-the-Art ML-Based Weather Forecaster

Explore modern machine-learning methods for producing useful, accurate forecasts from environmental data.

Conservation field research

Sustainability

Integrating ML into Conservation-Related Data Collection and Insight Generation

Apply machine learning to conservation data collection and turn field observations into actionable insight.

Attention mechanism research

Sustainability

Studying How Attention Mechanisms in LLMs can Lead to Innovative Climate Change Solutions

Study how attention-based models can support new approaches to climate research and decision-making.

Hands typing on a laptop keyboard

Bridging the Digital Divide

Accessible Software Evaluation

Audit software against accessibility standards and assistive-technology workflows, then publish barriers, severity, and recommended fixes.

A mentor helping adults learn to use laptops

Bridging the Digital Divide

Survey/Analysis of Technology Access in Communities

Study local barriers to devices, broadband, skills, trust, and language, then translate the findings into actionable recommendations.

Close-up of a circuit board and processor

Sustainability

Compressing Large Language Models

Investigate quantization, pruning, distillation, and other compression methods that make large language models smaller, faster, and less energy-intensive to run.

Close-up of a microprocessor on a circuit board

Sustainability

More Efficient ML Chips

Explore more sustainable GPU and machine-learning accelerator designs that reduce power use, heat, and material costs without sacrificing useful performance.

Start something

Have an idea that belongs here?

See how projects work