Machine Learning

Wrapping Up KALLM
The KANs are successfully reducing the number of weights required to achieve the same performance as the original model.
Last updated on Sep 3, 2025 Machine Learning
Midterm blog: CarbonCast Midpoint Update: From Vision to Reality
A midpoint update on the UCSC Open Source Research Experience (OSRE) project to build an API on top of CarbonCast, showcasing the enterprise-grade automation system and infrastructure built so far.
Last updated on Aug 3, 2025 web-development, Machine Learning, api
Midterm Report: KAN Integration into LLMs
Testing KANs in a language model to see if math-driven layers can do more with less.
Last updated on Jul 27, 2025 Machine Learning
CarbonCast
A UCSC Open Source Research Experience (OSRE) project to build an API on top of CarbonCast, enabling individuals and organizations to access and act on real-time carbon intensity forecasts.
Last updated on Jul 10, 2025 web-development, Machine Learning, api
CarbonCast
Kolmogorov-Arnold-based Transformer for LLMs
Developing efficient and interpretable LLMs by integrating Kolmogorov-Arnold-based modules into Transformers to reduce parameters and maintain performance.
Last updated on Jun 18, 2025 Machine Learning
Kolmogorov-Arnold-based Transformer for LLMs
LINQS: Autograder (LLM Detection)
This project aims to help academic institutions identify AI-generated code in student assignments, supporting fair grading and upholding academic integrity.
Last updated on Jun 15, 2025 Artificial Intelligence, Machine Learning, LLMs
LINQS: Autograder (LLM Detection)