Christopher Potts
@ChrisGPotts
Followers
14K
Following
2K
Media
116
Statuses
2K
Stanford Professor of Linguistics and, by courtesy, of Computer Science. Member of technical staff @stanfordnlp and @StanfordAILab. Co-founder @ Bigspin AI.
Joined November 2011
Every week brings a new agent framework or model. It's exciting, but founders need to know: what’s actually being used in production? We asked 413 senior builders how they’re adopting AI today, where adoption breaks, and where the real gaps are. Interactive findings ↓
6
6
31
Hat-tip to @vennemeyerd and colleagues for this insightful and timely work!
0
0
0
Just after the above clip, I consider the case of extreme sycophancy in 4o. Here's a lovely new paper applying mech-interp methods to the sycophancy problem:
arxiv.org
Large language models (LLMs) often exhibit sycophantic behaviors -- such as excessive agreement with or flattery of the user -- but it is unclear whether these behaviors arise from a single...
2
0
2
Interpretability research has made only minor contributions to AI safety so far. What can we do to change that? (Clip from a longer talk; YouTube link in the thread):
The Anthropic perspective on interpretability is prominent and significant, but not inevitable. My own take is quite different. (Clip from a talk I gave; YouTube link in the thread):
4
10
105
🌲 The Stanford School of Neural Network Interpretability 🌲
The Anthropic perspective on interpretability is prominent and significant, but not inevitable. My own take is quite different. (Clip from a talk I gave; YouTube link in the thread):
4
7
66
The Anthropic perspective on interpretability is prominent and significant, but not inevitable. My own take is quite different. (Clip from a talk I gave; YouTube link in the thread):
Severance as a show about interpretability research in AI (a clip from a talk; YouTube link just below):
3
25
198
I just learned that the one-of-a-kind Herumb Shandilya @krypticmouse, who is soon to graduate from Stanford with Master’s in CS, is on the job market for training infra, inference infra, or other ML engineering roles. Here’s a PSA to y’all: you should hire him.
6
9
51
We went from papers and no code, through PapersWithCode, to end up with ...
0
1
5
I'd also recommend working with @kmahowald if you can :) It doesn't just lead to award-winning papers, but it's always a great experience in itself!
If you are interested in winning an ACL paper award, it's a very smart move is to write something with @kmahowald. Historically, he has won at least one each year for as long as I can remember.
0
1
7
Severance as a show about interpretability research in AI (a clip from a talk; YouTube link just below):
7
26
217
I am delighted to be collaborating with both Alexa (@ARTartaglini) and Siri (@srihita_raju).
1
1
33
I'm recruiting PhD students this year with strong interest in NLP @UCLA! If you are interested in understanding how we use language to effectively communicate, and how those insights translate to NLP models, apply to join the @CoalasLab. Continue reading for details! (1/5)
11
100
529
If you are interested in winning an ACL paper award, it's a very smart move is to write something with @kmahowald. Historically, he has won at least one each year for as long as I can remember.
Delighted Sasha's work using mech interp to study complex syntax constructions won an Outstanding Paper Award at EMNLP! And delighted the ACL community continues to recognize unabashedly linguistic topics like filler-gaps, and the huge potential for LMs to inform such topics!
2
15
109
Delighted Sasha's work using mech interp to study complex syntax constructions won an Outstanding Paper Award at EMNLP! And delighted the ACL community continues to recognize unabashedly linguistic topics like filler-gaps, and the huge potential for LMs to inform such topics!
A key hypothesis in the history of linguistics is that different constructions share underlying structure. We take advantage of recent advances in mechanistic interpretability to test this hypothesis in Language Models. New work with @kmahowald and @ChrisGPotts! 🧵👇
1
22
90
Meet Slingshots // One. This inaugural batch includes leading-edge researchers advancing the science and practice of AI - with benchmarks, frameworks, and agents that ship real impact into the world. We're honored to support research from: @alexgshaw @Mike_A_Merrill
2
17
62
I am recruiting 2 PhD students to work on LM interpretability at UMD @umdcs starting in fall 2026! We are #3 in AI and #4 in NLP research on @CSrankings. Come join us in our lovely building just a few miles from Washington, D.C. Details in 🧵
14
171
776
In a last-minute change of events, I won’t be attending #EMNLP2025 in person. Still, I’m excited to share our poster for our paper, False Friends! https://t.co/gBFnoCbgRE
New paper! 🌈 In English, pie = 🥧. In Spanish, pie = 🦶. Multilingual tokenizers often share such overlapping tokens between languages. Do these “False Friends” hurt or help multilingual LMs? We find that overlap consistently improves transfer—even when it seems misleading. 🧵
0
5
51