Research

Publications

• Contrastive-Aware ViT for Weakly Supervised Semantic Segmentation

  Pranav Pativada, Zeeshan Hayder, Ali Zia, Nick Barnes

  IEEE Digital Image Computing: Techniques and Applications (DICTA), 2023
  Best Paper Runner Up Award
  Github

AI Safety

AI safety related research that I’ve done.

• Exploring the intersection of interpretability and optimisation

  AI Safety Fundamentals (Alignment) 2024

  Alignment Project Runner Up

  Github

• Refusal and Incapability as directions in LLMs

  An extension to Neel Nanda’s steering vector work

  Google Colab

P.S: These are also in my blog!

Papers

These are a collection of papers that I did as university research projects or extensions to my courses.

• Implementing Rabin’s Fingerprint Scheme for the Mojo Language

  2024 Semester 1 Advanced Studies Extension - Compilers - 91 HD

• Exploring the Effectiveness of Transformers for Long-Term Time Series Forecasting

  2024 Semester 1 Advanced Studies Extension - Statistical Machine Learning - 95 HD

• Solving Optimal Control Problems with Interior Point Dynamic Differential Programming

  2023 Semester 2 Advanced Studies Project - 97 HD

• Analysing the sequence diversity of secreted signal peptides in the Bacillus genus through novel machine learning techniques

  2022 Semester 1 Advanced Studies Extension - Genes and Replication - 85 HD

  Github

• Analysing lead carriers and examining their efficiency and properties in relation to dark matter studies using Accelerator Mass Spectrometry

  2021 Semester 1 Advanced Studies Project - 86 HD