Sitemap
A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.
Pages
Posts
A Reasoning First LLM Framework: How do train a LLM to reason?
Published:
P.S. This is my first draft. Article likely to go through more iterations and proofreading.
CPUs as a viable alternative to GPUs for Parameter-Efficient Fine-Tuning in Medical Summarization
Published:
CPUs as a Viable Alternative to GPUs for Parameter-Efficient Fine-Tuning in Medical Summarization
What is a Latent Space?
Published:
A Beginner’s Guide to Latent Space in Artificial Intelligence
portfolio
Medical Scribe
TLDR: An ambient AI assistant that frees clinicians from the documentation burden.
AI Search over Clinical Practice Guidelines
TLDR: Built a vertical search engine optimized for academic-style citations where doctors can query over 17 medical specialties and 122 Clinical Practice Guidelines in addition to the latest PubMed articles.
AI Symptom Checker
TLDR: Built a patient facing AI Symptom Checker that links patients to the appropriate level of care based on their symptoms 
AI Medical Coding
TLDR: Built an AI-powered medical coding engine that automates the translation of patient case details into ICD-11 codes, addressing the time-consuming, error-prone nature of manual coding with fast and scalable automation.
Domain Adaptation of Neural Multilingual Translation Models
TLDR: During my time at SIL Global, I researched low-resource machine translation focusing on data augmentation, data extraction, supervised fine-tuning, improving tokenizer support and growing embeddings to support new languages. The project is now maintained under Project Serval which integrates with Scripture Forge.
Semantic Entropy Probes
TLDR: To the best of my knowledge this is the first open-source implementation if Semantic Entropy Probes: Robust and Cheap Hallucination Detection in LLMs. This project aims to help researchers and developers identify hallucinations in large language models (LLMs) using an entropy-based method.
- Paper: Read the full paper on arXiv - GitHub Repository: Check out the code on GitHub
CNN Bone Age Classification
TLDR: Medical Imaging AI for Radiologists that helps predict Bone Age from Left Hand radiographs.
RSNA-MICCAI Brain Tumor Radiogenomic Classification
TLDR: Machine Learning solutions for brain tumor classifcation
Neural Machine Translation using HuggingFace, AWS and ClearML
A simple repository to perform NMT fine-tuning for machine translation
publications
Paper Title Number 3
Published in , 2015
The contents above will be part of a list of publications, if the user clicks the link for the publication than the contents of section will be rendered as a full page, allowing you to provide more information about the paper for the reader. When publications are displayed as a single page, the contents of the above “citation” field will automatically be included below this section in a smaller font.
Paper Title Number 4
Published in , 2024
The contents above will be part of a list of publications, if the user clicks the link for the publication than the contents of section will be rendered as a full page, allowing you to provide more information about the paper for the reader. When publications are displayed as a single page, the contents of the above “citation” field will automatically be included below this section in a smaller font.
Enhancing Medical Summarization with Parameter Efficient Fine Tuning on Local CPUs
Published in ICECCE. An IEEE Approved Conference Under Record # 63537, 2024
Documenting and summarizing patient symptoms and medical history for each visit can significantly burden clinicians’ time management. Large Language Models (LLMs) have demonstrated great potential in natural language processing (NLP) tasks; however, their effectiveness in clinical summarization tasks has not yet been rigorously validated. While much research has focused on leveraging closed LLMs like GPT-4, Claude, and Gemini for clinical applications, privacy concerns hinder their deployment in real clinical settings. On-premises deployment offers a potential solution. This study examines domain adaptation techniques on the open-source LLM, Llama 3 8B Instruct, to improve clinical summarization. Our approach emphasizes fine-tuning on CPUs instead of the more commonly used GPUs, aiming for greater cost savings in practical applications. We apply Quantized Low-Rank Adaptation (QLoRA) for efficient task-specific adaptation and introduce CPU optimization techniques such as IPEX-LLM and Intel® AMX to enhance performance. Our results show that CPU fine-tuning, while less conventional than GPU-based methods, still provides a practical, cost-effective, and privacy-aware solution for on-premises deployment, supporting the accuracy of medical summarization and enabling customization according to unique clinical requirements
Critique of Impure Reason: Unveiling the reasoning behaviour of medical Large Language Models
Published in Arvix, 2024
Background: Despite the current ubiquity of Large Language Models (LLMs) across the medical domain, there is a surprising lack of studies which address their reasoning behaviour. We emphasise the importance of understanding reasoning behaviour as opposed to high-level prediction accuracies, since it is equivalent to explainable AI (XAI) in this context. In particular, achieving XAI in medical LLMs used in the clinical domain will have a significant impact across the healthcare sector. Results: Therefore, we define the concept of reasoning behaviour in the specific context of medical LLMs. We then categorise and discuss the current state of the art of methods which evaluate reasoning behaviour in medical LLMs. Finally, we propose theoretical frameworks which can empower medical professionals or machine learning engineers to gain insight into the low-level reasoning operations of these previously obscure models. Conclusion: The subsequent increased transparency and trust in medical machine learning models by clinicians as well as patients will accelerate the integration, application as well as further development of medical AI for the healthcare system as a whole
talks
CPUs as a viable alternative to GPUs for PEFT in Medical Summarisation
Published:
Accelerating CPU based fine-tuning for LLMs to achieve GPU-like potential for Medical Summarization
teaching
Teaching experience 1
Undergraduate course, University 1, Department, 2014
This is a description of a teaching experience. You can use markdown like any other post.
Teaching experience 2
Workshop, University 1, Department, 2015
This is a description of a teaching experience. You can use markdown like any other post.