Hi, I'm Mahdi

@soda4fries

Welcome to my website

Welcome. I put things I find interesting here, if I need them later I can find them here. I think having a website is cool because everywhere else you cannot fully control how you present your information. Here you can arrange it however you want.

View My Blog
Profile photo

etc etc

Mostly House. Again.

also books →
  • Still from Breaking Bad

    Breaking Bad Say my name.

  • Still from Suits

    Suits I play the man.

  • Still from House M.D.

    House M.D. It's never lupus.

    Everybody lies.
  • Still from Better Call Saul

    Better Call Saul JMM: Justice Matters Most

    Your honor, s'all good, man!
  • Still from Lincoln Lawyer

    Lincoln Lawyer NTGUILTY

  • Still from Resident Alien

    Resident Alien Patience, Colorado

  • Still from Narcos

    Narcos Plata o plomo

Latest Blog Posts

Popular Tags

kubernetes (1) postgresql (1) database (1) devops (1) game (1) fun (1) interactive (1) markdown (1) formatting (1) guide (1) mdx (1) examples (1)
Setting up PostgreSQL on Kubernetes with CNPG

A guide to deploying and managing PostgreSQL clusters on Kubernetes using the Cloud Native PostgreSQL operator

kubernetespostgresql +2
Snake Playground

A classic Snake game I built for my company 404 page, then got addicted and added pathfinding algorithms.

gamefun +1
Blog System Formatting Guide

A showcase of all available markdown formatting features, MDX components, and interactive elements for this blog system.

markdownformatting +3
View All Posts

Papers

CVPR 2026 · Highlight

AutoDebias: Automated Framework for Debiasing Text-to-Image Models

AI image generators often pick up unwanted biases from training data — like always showing doctors as male. AutoDebias automatically hunts down these biases using vision-language models, then neutralizes them with CLIP-guided training, achieving 91.6% detection accuracy across 17 attack scenarios without breaking image quality.

arXiv:2508.00445

EMNLP 2025

Low-Confidence Gold: Refining Low-Confidence Samples for Efficient Instruction Tuning

Fine-tuning large language models requires massive datasets, but most of the data is redundant. Cluster training samples, score by confidence, and pick a curated 6K subset that outperforms models trained on the full dataset — saving huge compute costs.

arXiv:2502.18978

ICANN 2025

AgileIR: Memory-Efficient Group Shifted Windows Attention for Lightweight Image Restoration

Transformer models that restore blurry or degraded images eat enormous GPU memory. AgileIR redesigns the attention mechanism to work in groups across attention heads, cutting memory use by over 50% while keeping the same image quality (32.20 dB on Set5).

arXiv:2409.06206

IJCNN 2025

CFPFormer: Feature-pyramid like Transformer Decoder for Segmentation and Detection

Medical image segmentation needs models that see both fine-grained local details and big-picture context. CFPFormer combines feature pyramids with a Gaussian Attention mechanism in a transformer decoder, reaching 92.02% Dice Score on medical benchmarks — beating heavier ViT-based models.

arXiv:2404.15451