Research
At CTGT, we develop self-improving AI systems for high risk applications, and our research teams help us advance the state of the art in understanding the "black box" of AI.

A feature-level approach to mitigating bias and censorship in DeepSeek-R1
Cyril Gorlla


Empirical Analysis of Efficient Fine-Tuning Methods for Large Pre-Trained Language Models
Cyril Gorlla, Trevor Tuttle, Nigel Doering, Adhvaith Vijay


Training data eigenvector dynamics in the EigenPro implementation of the neural tangent kernel and recursive feature machines.
Cyril Gorlla
