Assistant Prof @columbia CS. Visiting Researcher @ Google DeepMind. PhD from @stanfordnlp. Language x Neural Nets.
Joined February 2015
- Tweets 244
- Following 56
- Followers 7,200
- Likes 488
51 Photos and videos
Pinned Tweet
Apr 29
New paper! Subliminal learning—transferring hidden signals between language models—is more powerful than we thought. By biasing the teacher with a steering vector instead of a prompt, we achieve strong, consistent transfer, which we use to study its mechanisms. w/@GeorgeMorgulis
6
35
302
20,034
Apr 29
New paper! Subliminal learning—transferring hidden signals between language models—is more powerful than we thought. By biasing the teacher with a steering vector instead of a prompt, we achieve strong, consistent transfer, which we use to study its mechanisms. w/@GeorgeMorgulis
6
35
302
20,034
Apr 29
This is the first paper from first author George Morgulis (@GeorgeMorgulis), a Columbia masters student. Many congratulations to him for getting this work out!
1
10
1,793
Apr 29
I’m most excited about how our subliminal steering enables the consistent study of subliminal learning across models and signals, and enabling future work in detection and mitigation.
paper: arxiv.org/pdf/2604.25783
github: github.com/GMorgulis/Sublimi…
1
19
1,752
Apr 23
I’m at both of today’s mentoring sessions! Come with questions; leave with either answers or more questions
Apr 23
ICLR office hour / mentoring sessions start today!
Walk-ins welcome.
⏰ Apr 23–25 12:00–12:45 and 1:15–2:00
📍 Rooms 206, 209, 212
If you're not sure where to go, just start at Room 206.
3
28
3,173
Apr 22
I’m at ICLR this year! Among other things, I’m happy to chat about PhD admissions; I’ll be hiring for my lab this upcoming cycle. Feel free to reach out.
1
5
151
14,752
John Hewitt retweeted
Apr 1
Excited to announce the 2026 iteration of the Communication & Intelligence Symposium at UChicago!
We have an amazing lineup of speakers @Diyi_Yang @johnhewtt @dashunwang @TomerUllman
We have a simple call for abstract that is due on Apr 15 (links 👇). Please come and share your research!
Co-organized with the awesome @universeinanegg and @divingwithorcas
2
15
77
35,131
Mar 4
Lots of interp thought discusses the linearity of the residual stream! This blog post: the residual stream isn't linear in a way that provides formal leverage, and interp methods based on linearity should not be preferred beyond empirical utility.
cs.columbia.edu/~johnhew/res…
5
17
235
13,079
Feb 26
In a pub trivia night, if you don't know the answer immediately, you "reason" through your memories -- is it X? no... was Y related?. In LMs, we find that code/math RLVR'd models' reasoning for this parametric knowledge access can be easily improved, say, by TriviaQA RLVR.
Feb 26
(1/8) Reasoning language models are great at math and code – but what about remembering facts stored in their parameters? Excited to share work with @johnhewtt exploring this!
TL;DR: we don't usually think of RLVR as useful for knowledge recall from parameters, but it helps a lot.
5
48
7,504
Jan 18
Hey folks, just in case it was unclear, I talked to Been and her account has been hacked, so please disregard.
3
2
51
17,304
9 Dec 2025
Excited to see more funding opportunities for interpretability, with an explicit call for strong evaluations.
7 Dec 2025
$1,000,000 to understand how LLMs write code.
Announcing: The Martian Interpretability Challenge.
Understanding the inner workings of LLMs is the greatest scientific challenge of our age,. Let's solve it.
Apply here: withmartian.com/prize
🧵👇
5
32
6,593
John Hewitt retweeted
21 Nov 2025
📢 Some big (& slightly belated) life updates!
1. I defended my PhD at MIT this summer! 🎓
2. I'm joining NYU as an Assistant Professor starting Fall 2026, with a joint appointment in Courant CS and the Center for Data Science. 🎉
🔬 My lab will focus on empirically studying the science of deep learning and applying deep learning to accelerate the natural sciences.
Very broadly interested in questions at the intersection of language, reasoning and sequential decision making. (Plus any other fun problems that catch our eye along the way!)
🚀 I am recruiting 2 PhD students for this cycle! If you're interested in joining, please apply here: cs.nyu.edu/dynamic/phd/admis… cds.nyu.edu/phd-admissions-r…
99
93
1,824
244,865
19 Nov 2025
Come do a PhD with me at Columbia!
My lab tackles basic problems in alignment, interpretability, safety, and capabilities of language systems. If you love adventuring in model internals and behaviors---to understand and improve---let's do it together!
pic: a run in central park
13
128
947
79,338
19 Nov 2025
I hire through the computer science department, and will be hiring 1-2ish PhD students this year. Columbia and New York have been an amazing place to live and do research.
And if you're not convinced, we just bought a mini fridge for snacks. Join us!
cs.columbia.edu/~johnhew/
5
12
106
19,032
23 Oct 2025
New work! Gemma3 can explain in English what it learned from data – when we distill that data into a new word (embedding) and query it for a description of the word. Gemma explained a word trained on incorrect answers as: “a lack of complete, coherent, or meaningful answers...”
4
30
191
36,786
23 Oct 2025
In one example, we taught Gemma a neologism that causes single-sentence answers. When asked for synonyms of this new word, it suggested “lack,” as in, “Give me a lack answer.” This didn’t look right, but indeed causes very curt answers. We call this a machine-only synonym.
1
9
2,001
23 Oct 2025
We see this as a step towards developing new language tools for learning about how language models store, process, and reason about potentially complex concepts—differently from how we do. Work with Oyvind Tafjord, Robert Geirhos, @_beenkim
Blog here: cs.columbia.edu/~johnhew//ne…
1
13
1,704