CS PhD candidate at @umdclip @umdcs advised by @psresnik working on understanding language. 2024, 2025 Summer @msftresearch, 2022 Summer @AdobeResearch
College Park, MD
Joined August 2016
- Tweets154
- Following273
- Followers241
- Likes1.6K
Pinned Tweet
Linguistic theory tells us that common ground is essential to conversational success. But to what extent is it essential? Can LLMs detect when humans lose common ground in conversation?
Our ACL 2025 (Oral) paper explores these questions on real-world data.
#ACL2025NLP #ACL2025
Rupak retweeted
engineering”. But models are post-trained to “solve” tasks and maximise their reward. So we try to capture the abstract domain in a (scalable, verifiable) set of problems.
No matter how much we try, when models are trained to solve these problems, we realise that some parts of
Rupak retweeted
His eyes were a striking blue, like blue Gatorade, and his cheeks were flushed, like red Gatorade. His demeanor was mysterious, yet refreshing, like yellow Gatorade. And his shirt was the hue of orange Gatorade
Rupak retweeted
TMLR has faced a deluge of submissions, necessitating stricter desk rejection policies due to limited reviewer capacity
Co-EiC Nihar Shah reached out to authors of 10 papers slated for desk reject. Could they answer questions about their *own* submission?
medium.com/@TmlrOrg/asking-a…
Rupak retweeted
I think breaking down mathematicians' motivations into like 'wanting to understand math' and 'ego' misses a lot. Ppl value participating in a process where the growth of their individual understanding has a rich two-way relationship to the growth of collective understanding
Rupak retweeted
Come join us in Boulder! Feel free to get in touch if you have any questions.
jobs.colorado.edu/jobs/JobDe…
Rupak retweeted
but they were, all of them, deceived, for another watch was made…
Sam Altman @sama just ordered 7 ultra-rare custom Swiss watches from Vanguart, fully branded with the OpenAI logo on the dial. One for him, six for his inner circle.
Inscribed on the back of each watch is ‘if AGI.aligned: deploy()’ which was written in the programming language Python.VANGUART
This is starting to be a trend in the tech world and I’m all for it
—
If your company wants customer branded Rolex or other watches shoot me a dm we specialize in bulk corporate orders with custom engraving and dial branding
Rupak retweeted
Wherever you live in the USA, in-state tuition at your state’s flagship university is an incredible bargain.
Kids: Don’t overthink college. Don’t stress about getting into the Ivies. Go to your state school, go to some football games, take the toughest classes in a serious major. Take five years if you need to work an occasional semester to keep the debt down. You’ll be fine.
My daughter's tuition bill just arrived: $4,400 for the semester at UNC.
Last year, it was $44,000 at a different school.
Same kid. Same major. Ten times the price.
She started at a private college out of state. Great school. But the department she cared about wasn't especially strong. The flights were miserable. Every semester cost as much as a new car. And she was homesick.
This year she transferred to UNC, which she'd originally gotten into and turned down. (Because what 18-year-old picks the school Dad wants?)
In-state tuition at one of the best public universities in the country turns out to be less than a lot of families spend on summer camp.
A friend put it well: one of the best things about North Carolina is that there's a college for almost anyone willing to put in the work. From essentially open-enrollment schools to UNC and Duke, you can find it without leaving the state.
Like most parents, I figured the price tag said something about the quality.
Then I watched my own kid find an equal or better education for one-tenth the cost, 20 minutes from home.
Rupak retweeted
The idea that you can save time on writing to focus on ideas is like saying you can save time on lifting weights to focus on building muscle.
Rupak retweeted
🧵 Trusting AI is not one decision.
In our ACL 2026 paper, expert trivia players had to decide when to let AI answer, when to ignore it, and when to change their minds because of it.
The main failure mode was not blind trust.
Rupak retweeted
Unfortunately for @WSJopinion, @nytopinion, and @PostOpinions, slop-eds continue to rise. In June, roughly 6.5% of opinions are flagged for AI-use.
Rupak retweeted
The topic for the PhD is open, so a genuine intellectual curiosity is the main criterion
I have one open position now (deadline July 9); another this fall (see next post). Apply!
* alexanderhoyle.com/posts/wor…
* jobs.tuwien.ac.at/Job/266689 (a nice bonus: pay exceeds top US programs)
Rupak retweeted
New preprint out!
Recent work tasks LLM agents with re-running existing social science replication code. Given current capabilities, that should be table stakes
Here, we move up a level of abstraction, and ask models to reproduce results from a paper’s descriptions alone
Can AI agents read a social science paper and write the code from scratch to reproduce its results?
No access to original code. Just text + data.
New paper with Ben Kohler, @david_rzs, @__jae_1, and @miserlis_
👇
Rupak retweeted
Is anyone in my network looking to hire an RA for 1 year? I know a masters grad from the American University of Beirut with 2+ years of exp in Health AI and pubs in bioinformatics. Has a PhD offer but with funding constraints can’t start, so waiting until next cycle to reapply!
Rupak retweeted
🔊 @UMDCS Assistant Professor @sarahwiegreffe discusses her path into computer science, research on interpreting large language models and advice for students entering #AI research in a new Q&A.
Read more: go.umd.edu/Wiegreffe-2-2026
Rupak retweeted
#HBD to arXiv!🎈
On August 14, 1991, the very first paper was submitted to arXiv. That's 34 years of sharing research quickly, freely & openly!
Some baby pictures to show how far we've come . . . when we were just a computer under desk . . . & in our 1994 punk phase . . . 👶💾
Rupak retweeted
🚀New dataset release: WildChat-4.8M
4.8M real user-ChatGPT conversations collected from our public chatbots:
- 122K from reasoning models (o1-preview, o1-mini): represent real uses in the wild and very costly to collect
- 2.5M from GPT-4o
🔗 hf.co/datasets/allenai/WildC… (1/4)
Thrilled to see WildChat featured by @_akhaliq, just as predicted by AKSelectionPredictor!😊
Explore 1 million user-ChatGPT conversations, plus details like country, state, timestamp, hashed IP, and request headers here:
huggingface.co/datasets/alle…
Rupak retweeted
When questions are poorly posed, how do humans vs. models handle them? Our #ACL2025 paper explores this + introduces a framework for detecting and analyzing poorly-posed information-seeking questions!
Joint work with @boydgraber & @rachelrudinger!
🔗 aclanthology.org/2025.acl-lo…
