TWC Undergraduate Research Program
As part of the university-wide URP program, TWC provides full-time undergraduate students the opportunity to pursue research projects and work directly with faculty members on real-world, hands-on projects related to the above topics. Students may pursue research for academic credit or receive funding (when available) for their project. An undergraduate research laboratory at TWC is available to accepted students, which includes private study locations, close proximity to world-leading faculty, state-of-the-art technology, as well as other benefits. Program openings are announced in the beginning of the fall and spring semesters.
A TWC URP/Internship Application Form is available for students who are interested in the program.
Please refer to Rensselaer's Undergraduate Research website for more information.
Highlights of Available Projects:
The EFEMO Web Tools project is a web-based platform for exploring existing explanation faithfulness measures, represented as an ontology-driven knowledge graph, and adding new measures to the knowledge graph. This will be developed as part of the Explanation Faithfulness Evaluation Measures Ontology project at TWC which models different methods to determine if an explanation is representative of an AI model's decision-making process.
The goal of the project is to develop a web-based tool that allows non-experts to explore and add to the knowledge graph. The tool must be easy to use and maintain performance as the graph grows.
Neurosymbolic AI Research: We are looking for an undergraduate with strong programming skills to assist with a neurosymbolic AI research project. The goal of the project is to develop infrastructure and tools to enhance the reliability of generative AI models by validating model-generated outputs against symbolic rules and constraints specified in an ontology. The student's role will be to assist with the implementation of increased functionality for the system, development of a test suite, and gathering data to evaluate the system's performance. The project involves use of Python, CUDA, ontologies, scripting, and local LLMs; background in these areas is helpful, but not required.
In particular, we are looking for a student with experience in the following areas:
- Writing programmer-friendly code, with well-made APIs and documentation
- Gathering evaluation data and performing data analysis
- Understanding of algorithms and time complexityThe Idea Graph Website Project is a web-based platform for exploring and visualizing scholarly knowledge represented as a multi-layered, schema-driven knowledge graph. Developed as part of the Idea Graph project at TWC, the system models research artifacts across multiple abstraction layers, from document metadata and textual structure, to named entities, and further to high-level semantic constructs such as research problems, claims, evidence, and arguments.
The platform is grounded in the Idea Ontology, which provides a discourse-aware schema for linking document structure, semantic entities, argumentation roles, provenance, and versioning information. A central objective is to deliver an interactive web UI that enables users to navigate large research collections, inspect fine-grained semantic annotations, and trace the evolution of ideas across the scholarly ecosystem.