About Me
This is the website of Sidharth Pulipaka. I’m an undergraduate student, grinding Math/Deep Learning. It’s been a little over two years since I started my grind.
I believe AGI is almost here (Kaplan et al. [2020], Hoffmann et al. [2022]).
Neural Scaling Laws are like sheet music, the important thing isn’t can you read music, it’s can you hear it. Can you hear the music, Robert?
I think of AGI as something that can potentially replace almost all the jobs. Post-AGI world is going to be so different, and we are so early. Meanwhile, I’m spending all my time on Deep Learning hoping to contribute my worth’s work in achieving safe AGI.
Let’s have a thought experiment. Let’s say we achieve AGI. It’s completely aligned. It’s open-source if you’d like it that way. All the governments of the world are at UN General Assembly, and have decided to completely co-operate with one another. Everyone trusts one another and no one is racing for global dominance. What path will humanity take going forward? How would such a future look like?
I think about this a lot.
I’d like to be a good AI Alignment Researcher one day, but automated Alec Radford is almost around the corner. But meanwhile, here’s my experience for my future employers (and also just in case AGI is far away):
Experience
Research Intern at AI4Bharat @ IIT Madras (January 2025 - Present)
Paper Breakdowns
If you are on this website, that’s probably for this.
Check out Paper Breakdowns where I upload a breakdown of an interesting paper (usually the latest as soon as they come out) very frequently. I try to avoid making posts on popular papers rather focus on making posts on niche papers.
Goal is not to teach beginners (there are already so many on the internet). My breakdowns will usually be high-level, and I assume the reader knows their basics well. I would really like to read posts which give a quick and fairly in-depth overview of papers, but I did not find any on the internet that I liked (there are quite a few YouTube channels though). I could not find any such blogs, so I decided to maintain one.
WARNING: I’m not an expert, and I’m just learning. Let me know if I make any errors anywhere.
In the days of AI slop, spelling and grammatical errors are watermarks of human generated content.
Update: alphaxiv. has an amazing feature that lets us turn any paper into a blog. It is a game-changer. Those are the exact kind of blog-posts I was hopig for. There is no point for me to do paper-breakdowns when AI can do it in an instant. I got replaced by AI already.
Publications
Pulipaka, S., et al. “Mark My Words: A Robust Multilingual Model for Punctuation in Text and Speech Transcripts.”
arXiv Preprint: arXiv:2506.03793. Under review as a conference paper at AACL (2025)
Contact Me
You can contact me via X (formerly Twitter) where you can find me as @psidharth567
Or you can email me at psidharth567@gmail.com