UROP Project
AI That Remembers: Making Language Models Faster by Recycling Their Memory
LLM, KV cache, AI efficiency, ML systems
Research Mentor: Oteo Mamo,
Department, College, Affiliation: Computer Science, Arts and Sciences
Contact Email: om21d@fsu.edu
Research Assistant Supervisor (if different from mentor):
Research Assistant Supervisor Email:
Faculty Collaborators:
Faculty Collaborators Email:
Department, College, Affiliation: Computer Science, Arts and Sciences
Contact Email: om21d@fsu.edu
Research Assistant Supervisor (if different from mentor):
Research Assistant Supervisor Email:
Faculty Collaborators:
Faculty Collaborators Email:
Looking for Research Assistants: Yes
Number of Research Assistants: 2
Relevant Majors: Computer Science, Computer Engineering, Data Science, Applied Mathematics, Statistics, or any major with Python experience and curiosity about AI
Project Location: On FSU Main Campus
Research Assistant Transportation Required: Yes Remote or In-person: Partially Remote
Approximate Weekly Hours: 10, Flexible schedule (Combination of business and outside of business. TBD between student and research mentor.)
Roundtable Times and Zoom Link:
Number of Research Assistants: 2
Relevant Majors: Computer Science, Computer Engineering, Data Science, Applied Mathematics, Statistics, or any major with Python experience and curiosity about AI
Project Location: On FSU Main Campus
Research Assistant Transportation Required: Yes Remote or In-person: Partially Remote
Approximate Weekly Hours: 10, Flexible schedule (Combination of business and outside of business. TBD between student and research mentor.)
Roundtable Times and Zoom Link:
- Day: Tuesday, September 1
Start Time: 12:00
End Time: 2:00
Zoom Link: https://fsu.zoom.us/my/om9740
Project Description
Every time an AI model like ChatGPT reads a document, it builds an internal "memory" of what it read, and today that memory is constantly thrown away: when the model is updated, when the conversation ends, when space runs out. Rebuilding it from scratch wastes enormous amounts of GPU time, money, and energy. Our group develops methods that let models keep, compress, and reuse this memory, and we release them as open-source tools (SALT - accepted at EMNLP 2026, one of the top AI conferences). As a research assistant you'll run real experiments on GPUs, analyze the results, and contribute to open-source software. No prior AI experience required.Research Tasks: Literature review of recent AI efficiency papers.
Running experiments on GPU servers, data analysis and visualization in Python, and reproducing results from published papers.
Testing, documentation, and issue triage for our open-source repositories (SALT, CRELAY).
Skills that research assistant(s) may need: Required: basic Python, curiosity.
Recommended: Git/GitHub, Linux command line, NumPy or PyTorch.