Welcome to my GitHub profile! I'm a PhD student at Imperial College London, with a focus on Large Language Models. My research aims to improve LLMs by identifying their limitations and enhancing their understanding capabilities.
-
Scientific Language Models
- Developing models tailored for scientific literature and discourse.
- Enhancing the accuracy and efficiency of LLMs in understanding and generating scientific content.
-
Factually Correct Fair Representations
- Ensuring LLMs generate factually accurate and unbiased outputs.
- Researching methodologies to mitigate biases and enhance fairness in model predictions.
-
Applied Scientist Intern @ Amazon [October 2024 - Ongoing]
- Working on innovating machine translation methods.
-
AI Research Intern @ IBM Research [Summer 2024]
- Explored the knowledge and language quality gaps in debiased language models. Investigated faithful and fair language generation methods.
-
AI Research Intern @ IBM Research [Summer 2023]
- Developed a computational feedback based fine-tuning method for harnessing generative capabilities of large language models on sensitive downstream tasks where human evaluations are ambigious and expensive to obtain.
-
Machine Learning Engineer II @ Comcast NBCUniversal
- Implemented machine learning solutions to improve the forecasting models for click-rate, video-completion rate and as such metrics.
- End-to-End Large Models Training
- Investigating the secrets of efficient training processes for large models.
- Sharing insights and methodologies to contribute to public knowledge in this domain. An outdated but quite detailed chronicle for 175B model training
- Favourites repos so far: GPT with C/CUDA , Llama3 from scratch