Skip to content
View msmrexe's full-sized avatar
☄️
lost in outer space
☄️
lost in outer space

Organizations

@PsyCity @VAST-GAME @TGML-Lab @MsMaryamRezaee @SUT-CS-GM

Block or report msmrexe

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
msmrexe/README.md

🐳 From Mer to Earth: Greetings, earthling!

MR Banner

I'm a computer science researcher specializing in LLM Interpretability and Cognitively-Inspired AI. I'm passionate about building models that are not only powerful but also logically understandable. Though I should add, I do tend to diverge and dabble in a quite a few more passion projects or fascinating alogithmic questions as well. :)

Current Highlights of My Life

  • 🔎 Finishing my Master's thesis on generative model interpretability (the repository is still private, whoops).
  • 🧠 Leading the design of an extremely fun and challenging educational C.S. game for bachelor students.
  • 🔭 Exploring concept-based interpretability methods for RAG systems as a research mentor.
  • 🤖 And of course, beginning my journey into Neurosymbolic AI for my future PhD.

About Me

My core interest lies at the intersection of machine learning and human cognition; I've been diving deeply into both, and mean to merge conceptual models of the human cognition with the design of ML systems to improve how models think and hopefully perform. When I'm not working on research, I enjoy creative coding, developing tools for other developers (like VSCode themes), and applying mathematical concepts to practical programming challenges.

🛠️ My Tech Stack

⠀AI & Machine Learning

Python PyTorch TensorFlow Hugging Face Scikit-learn NumPy Pandas

⠀Other Languages & Databases

JavaScript R MATLAB Bash RISC-V PostgreSQL

⠀Design & Illustration

Illustrator InDesign Photoshop Figma HTML5 CSS3

⠀Tools & Platforms

LaTeX Git GitHub VSCode

My GitHub Stats


📫 Get In Touch

I'm always excited to connect with fellow researchers, developers, and creative minds. Whether you have a question about my work, spot a fascinating problem, or just want to discuss the future of AI and cognition, please don't hesitate to reach out. You can find me here:

Pinned Loading

  1. violet-night-vscode-theme violet-night-vscode-theme Public

    A sleek, dark, and vibrant VSCode theme with violet accents; easy on the eyes and perfect for night coding.

    2

  2. neurosymbolic-vqa-program-generator neurosymbolic-vqa-program-generator Public

    A comprehensive implementation of a Neurosymbolic framework for Visual Question Answering (VQA) on the CLEVR dataset. This project translates natural language questions into symbolic programs using…

    Python 1

  3. llm-math-reasoning-analysis llm-math-reasoning-analysis Public

    An evaluation of prompting techniques (Zero-Shot CoT, Few-Shot, Self-Consistency) on the Mistral-7B model for mathematical reasoning. This project systematically benchmarks 7 distinct methods on th…

    Python 1

  4. pytorch-pix2pix-image-translation pytorch-pix2pix-image-translation Public

    A PyTorch implementation of the Pix2Pix model (cGAN) for image-to-image translation. This project translates semantic segmentation maps to photorealistic images using the Cityscapes dataset.

    Python 1

  5. pytorch-lora-from-scratch pytorch-lora-from-scratch Public

    A from-scratch PyTorch implementation of Low-Rank Adaptation (LoRA) to efficiently fine-tune BERT models for text classification. This project compares the performance and parameter efficiency of L…

    Python 2

  6. pytorch-gpt2-persian-sentiment-generation pytorch-gpt2-persian-sentiment-generation Public

    A from-scratch implementation of a scaled-down GPT-2 model in PyTorch, trained on the Snappfood dataset for sentiment-controlled Persian text generation.

    Python 1