Welcome to the Blog

Welcome! This is the first post on my blog — a place where I think out loud about my research and share work in progress. To add a new post, I just drop a Markdown file into the posts/ folder and push; the site rebuilds and the new post appears automatically.

What I’ll write about

My research sits at the intersection of graph learning and natural language processing, with a growing focus on self-improving agentic systems. Expect posts on:

• Agentic AI: self-improving and evolving multi-agent systems
• Retrieval-Augmented Generation (RAG) over graph-structured knowledge
• LLM post-training: distillation and self-supervised objectives
• Spatio-temporal and graph representation learning

Writing posts

Each post is plain Markdown with a small header at the top:

---
title: My Post Title
date: 2026-07-01
summary: A one-line description shown on the blog index.
tags: [RAG, Graphs]
---

Everything below the header is standard Markdown. Code blocks are syntax highlighted:

def attention(q, k, v):
    scores = q @ k.T / (k.shape[-1] ** 0.5)
    weights = softmax(scores, axis=-1)
    return weights @ v

Tables work too:

And blockquotes render with a clean left border — handy for highlighting a key idea or a quote from a paper.

Versioned, like a paper

Each post keeps a version history, derived from its git commits — much like arXiv. You can browse any past version and use Compare versions to see a line-by-line diff of how a post (and the thinking behind it) evolved over time.

That’s it. Thanks for reading!