2025 Annual Summary — Growth, Challenges & Looking Ahead

“The measure of intelligence is the ability to change.” — Albert Einstein

Another year has passed. 2025 was a year of unexpected turns, quiet growth, and lessons learned the hard way. This is my attempt to look back honestly before stepping into what comes next. 🌅

📚 Academic & Research

Papers & Projects

This year I worked on several exciting problems at the intersection of numerical analysis and machine learning. Key highlights:

• 📄 Submitted my first paper on tensor train approximations — still under review, fingers crossed 🤞
• 🔬 Started exploring physics-informed neural networks (PINNs) as a side project
• 💡 Read ~40 papers (finally started using Zotero properly)

My usual 2am research setup. Coffee not optional.

What I Learned

The biggest technical insight this year was realizing that algorithmic elegance and practical efficiency are often in tension. A method that looks beautiful on paper can be completely impractical at scale — and vice versa. Learning to navigate that gap has been humbling.

🌏 Life & Experience

Places I Went

Moments Worth Remembering 📷

One of those rare evenings where everything just felt right.

Some smaller moments that I don’t want to forget:

1. The afternoon I finally understood the proof I’d been stuck on for three weeks
2. A long walk along the Schuylkill River after a rough seminar
3. Cooking a proper meal for the first time in months and realizing I’d missed it

💪 Health & Habits

What Stuck

• 🏃 Running: managed 3× per week for most of the year — proud of this one
• 📖 Reading before bed instead of scrolling (works maybe 60% of the time)
• 🧘 Started a very modest meditation habit (5 min/day, nothing heroic)

What Didn’t

• ❌ Sleep schedule — still a disaster
• ❌ “I’ll reply to that email tomorrow” — I did not reply to that email
• ❌ The guitar I bought in January has been used approximately twice

🧠 Things I Changed My Mind About

Updating your beliefs when evidence changes is a feature, not a bug.

• On productivity systems: Elaborate setups don’t work for me. A plain text file and a pen do.
• On asking for help: I used to see it as a weakness. Now I think it’s just efficient.
• On “staying busy”: Busy ≠ meaningful. This one is still sinking in.

📊 By the Numbers

Papers read:         ~40
Code commits:        312
Books finished:       7
Countries visited:    2
Times I said "almost done" on a project that wasn't:  countless

🔭 Looking Ahead — 2026 Goals

These aren’t resolutions. They’re directions:

• Submit at least 2 papers
• Build something (anything) that other people actually use
• Travel somewhere purely for the experience, not work
• Sleep before 1am at least half the time
• Call family more

🙏 Gratitude

None of this year happened in isolation. Thanks to my advisor for the patience and the hard questions. To friends who showed up. To strangers on the internet whose code I used and whose blogs I read at 3am when I was stuck.

This is the part I find hardest to write, not because I’m not grateful — I am — but because gratitude is easier to feel than to express. So I’ll just say: thank you.

See you in 2026. 🎆

Posted on December 31, 2025 · Reading time ~4 min