Soham De

Major personal update: After a bit of hard work and a whole lot of extreme good luck, I'm happy to share that I'll be joining @UW (@uw_ischool) to start my PhD this Fall! It has been an extremely rewarding journey so far and I look forward to the next phase!

Could social media make us less polarized instead of more? We tested 5 algorithms on 3 platforms with 10,000 people for 6 months during the 2024 election, and found that the answer is yes. 🧵

🚨🎉Excited to announce that our paper “Grok in the Wild: Characterizing the Roles and Uses of Large Language Models on Social Media” is accepted at @icwsm 2026! In this paper, we investigate how, when, and to what effect Grok is used on X. 🧵1/n thread

@ma_tay_ Congrats Dr Sorensen! 🎉

🎓🎓🎓!!!

@agdhruv I had no idea you had made what we used to call MessCat haha 🤣 I remember maintaining it for a while and I wonder if it’s still in use?

An Ode to Coding Before AI Like most others, I learned to code in 2014, long before AI coding agents, autocomplete, and such. My friend and I would exchange HTML/CSS/JS tutorials on 1TB hard drives. I’d flip between VLC, Sublime Text, and Chrome on my family’s old HP laptop, trying to center a div or fix a PHP bug I barely understood. We built websites at high school tech competitions without any internet access. In college, I built a lowkey viral Facebook Messenger chatbot that shared our campus food menu and bus timings. It crashed when someone sent it a cat GIF haha. I "vibe-coded" so many other projects (the vibe came from my inexperience) and encountered creative bugs on the way. Every bug was painful, but every bug taught me something: OS quirks, file systems, logs, APIs, threading, the GIL, tensor dimensions. These days, I use Cursor, Claude Code, and all the AI tools. They’re amazing! But I do feel a weird nostalgia. When AI writes so much of the code, you enjoy the product more than the process. As an engineer, I was wired to enjoy the process, but that's changing. AI coding makes me productive, but a part of me misses when coding was slow, messy, and deeply personal...

@ma_tay_ @humansand Congratulations!!

Also happy to share that I've joined @humansand!! More details in the coming days, but I am SO excited to build AI to work with and be good for humans!

All thanks to my amazing collaborators @Effeietsanders @ashishgoel @SmithaMilli @ArielProcaccia and Alice Siu!!! This was some work I did @AIatMeta this summer (more on the way!) Read more here: arxiv.org/abs/2511.04588 (7/7)

Our auditing tools are integrated into a deliberation platform @StanfordDDL used in 50+ countries, allowing moderators to measure and improve representation in real time (thanks, Harshvardhan Agarwal and others!) (6/7)

🚨 New paper! 🔍 “Question the Questions: Auditing Representation in Online Deliberative Processes” In deliberative polls, participants propose questions for experts but only a few make it to the panel. How representative are those chosen questions of everyone’s interests? 🧵👇

(please reshare) I'm recruiting multiple PhD students and Postdocs @uwcse @uwnlp (bdata.uw.edu). Focus areas incl. psychosocial AI simulation and safety, Human-AI collaboration. PhD: cs.washington.edu/academics/phd/… Postdocs: docs.google.com/document/d/1h4…

Today we're releasing Community Alignment - the largest open-source dataset of human preferences for LLMs, containing ~200k comparisons from >3000 annotators in 5 countries / languages! There was a lot of research that went into this... 🧵

@hungrysohan some have it worse …

not a great day to be named Sohan

@shahanmemon Hahaha the chances of me being employed by a YC startup were always slim but are finally 0 😂

@actuallysoham Lol was about to tag you in some related tweets!

interesting day on the internet for @actuallysoham 🙃

@Marco_Piani Your evaluation idea definitely makes sense to compare AI-assisted vs human written notes (we do something similar here: dl.acm.org/doi/10.1145/36…) but not sure if that directly solves the bottleneck issue?

@Marco_Piani Yea, I agree with rating being a bottleneck and I think there's a lot of interesting ways to address that issue. For instance, tools/algorithms to assist human raters, transfer ratings across similar notes etc.

Here’s our articulation of what the future of Community Notes in an LLM era could look like 👀 : Very interested to hear your thoughts!