Emmy Liu @_emliu

PhD student @LTIatCMU, working with @gneubig on NLP || intern @AIatMeta || UofT ‘21 🇨🇦 ||🤖✨🔡 nightingal3.github.io Pittsburgh, PA Joined November 2021

Tweets

289
Followers

1K
Following

510
Likes

1K

CLS @ChengleiSi

2 months ago

Are AI scientists already better than human researchers? We recruited 43 PhD students to spend 3 months executing research ideas proposed by an LLM agent vs human experts. Main finding: LLM ideas result in worse projects than human ideas.

11 181 624 145K 215

Download Image

Kiril Gashteovski @kgashteo

2 weeks ago

Happy to share that our paper got accepted at #EMNLP2025! In this work, we study which factors—other than scaling model size or training data size—affect the performance of LLMs. For details, see Emmy's thread 👇

Emmy Liu @_emliu

6 months ago

5 56 214 19K 133

Download Image

0 1 5 571 1

Emmy Liu @_emliu

2 weeks ago

Check out this exciting workshop at CMU on Sept 12! AI for science is an exciting upcoming area, hope to hear some great discussions!

Jiayi Geng @JiayiiGeng

2 weeks ago

Check out this exciting workshop at CMU on Sept 12! AI for science is an exciting upcoming area, hope to hear some great discussions!

1 13 127 25K 35

0 0 10 659 1

Hokin Deng @DengHokin

a month ago

#embodied All forms of biological intelligence are grounded movements🏃‍♂️ muscles & motor neurons 🧠 emerge before visual cortex & rods & cones in eyes 👁️ Building monocular better-than-mocap-studio #video2motion is our critical step towards human embodied intelligence.

myolab.ai @MyolabAI

a month ago

0 18 31 7K 14

Download Video

1 19 32 4K 11

Graham Neubig @gneubig

4 months ago

I do think that AI has a lot of promise for science, but we need lots of serious work to get there! One thing I'm very interested in is how we can build AI systems that can effectively judge the quality of research.

0 1 16 2K 0

Emmy Liu @_emliu

4 months ago

Returning from #NAACL2025 & had some interesting discussions! One topic that came up a lot was AI scientists and how they should be implemented, evaluated, etc. Used this as inspiration to finish up a blog post on AI science and the state of reviewing: nightingal3.github.io/blog/2025/04/2…

4 7 82 10K 22

Kayo Yin @kayo_yin

6 months ago

Induction heads are commonly associated with in-context learning, but are they the primary driver of ICL at scale? We find that recently discovered "function vector" heads, which encode the ICL task, are the actual primary drivers of few-shot ICL. arxiv.org/abs/2502.14010 🧵

16 112 779 277K 663

Download Image

Emmy Liu @_emliu

4 months ago

Join us this week at the #NAACL student research workshop for an exciting keynote presentation by @psresnik! Fun fact, he was a co-creator of the SRW back in the day, so this is coming back full circle ⭐ ⏱️ May 1st, 4-5PM MDT 📍 San Miguel, or online

0 4 15 2K 0

Download Image

Belinda Li @belindazli

6 months ago

Past work has shown that world state is linearly decodable from LMs trained on text and games like Othello. But how do LMs *compute* these states? We investigate state tracking using permutation composition as a model problem, and discover interpretable, controllable procedures🧵