Lightnews — Scholar-powered news

Najoung Kim @najoung.bsky.social · Jul 25

Required qualifications: BA, hands-on engineering experience, and interest in research. Please help spread the word and share with people who might be a good fit! 🪄

Unfortunately only open to US citizens/permanent residents/independent work auth holders.

Najoung Kim @najoung.bsky.social · Jul 25

The first part of this effort was RExBench: bsky.app/profile/najo...

There will be some freedom of scope for the second part of the project within the theme of research agent evaluation - the RA will contribute to scoping the project along with the team as well.

Najoung Kim @najoung.bsky.social · Jul 2

Seeing an experiment and thinking "but have they tried X? what if we do Y?" is a key part of research and a start to new discoveries. RExBench tests if coding agents can implement new extensions.

It complements recent evals (eg PaperBench from OpenAI
) on replication! See 👇 for details

Sebastian Schuster @sebschu.bsky.social · Jul 2

Can coding agents autonomously implement AI research extensions?

We introduce RExBench, a benchmark that tests if a coding agent can implement a novel experiment based on existing research and code.

Finding: Most agents we tested had a low success rate, but there is promise!

1

Najoung Kim @najoung.bsky.social · Jul 25

👾 Full-time research assistant position (1 year) with @sebschu.bsky.social and me! 👾

We're looking for someone to join the research agent evaluation team, starting Fall 2025. Application link to be available soon, but feel free to send us your CV and/or come talk to us at #ACL2025. 🧵

1 4 10

Reposted by Najoung Kim

Lizzie Coppock @lizzieloo.bsky.social · Jul 24

I have a sabbatical coming up and I'm going to Nepal! Why Nepal? You can read about it at my first blog post about this trip.

TL;DR: linguistic diversity, writing systems (and pretty scripts), classifiers, and a school for Newar kids in Kathmandu.

sites.bu.edu/lislab/2025/...

Field trip to Nepal! | Linguistic Semantics Lab (LiSLab)

sites.bu.edu

1 7

Najoung Kim @najoung.bsky.social · Jul 22

ever since VLMs were a thing i've been interested in how the additional visual modality changes language in meaningful ways. after negative findings after negative findings, excited to report this result! proud of our junior authors for digging into this 🐸

yuluqin.bsky.social @yuluqin.bsky.social · Jul 22

Does vision training change how language is represented and used in meaningful ways?🤔The answer is a nuanced yes! Comparing VLM-LM minimal pairs, we find that while the taxonomic organization of the lexicon is similar, VLMs are better at _deploying_ this knowledge. [1/9]

3 12

Najoung Kim @najoung.bsky.social · Jul 8

green carded finally 💚💚

9

Najoung Kim @najoung.bsky.social · Jul 2

Excited this is finally out and ❤️ to our team @nedwards99.bsky.social @yukyunglee.bsky.social Audrey Mao and Yulu Qin! of course @sebschu great as always 🙏🙏

1

Najoung Kim @najoung.bsky.social · Jul 2

Seeing an experiment and thinking "but have they tried X? what if we do Y?" is a key part of research and a start to new discoveries. RExBench tests if coding agents can implement new extensions.

It complements recent evals (eg PaperBench from OpenAI
) on replication! See 👇 for details

Sebastian Schuster @sebschu.bsky.social · Jul 2

Can coding agents autonomously implement AI research extensions?

We introduce RExBench, a benchmark that tests if a coding agent can implement a novel experiment based on existing research and code.

Finding: Most agents we tested had a low success rate, but there is promise!

Screenshot of the RExBench preprint title page.

1 4

Reposted by Najoung Kim

Koyena Pal @koyena.bsky.social · Jun 30

🚨 Registration is live! 🚨

The New England Mechanistic Interpretability (NEMI) Workshop is happening Aug 22nd 2025 at Northeastern University!

A chance for the mech interp community to nerd out on how models really work 🧠🤖

🌐 Info: nemiconf.github.io/summer25/
📝 Register: forms.gle/v4kJCweE3UUH...

8 10

Najoung Kim @najoung.bsky.social · Jun 18

i'll be in copenhagen for a few days, lmk if you want to get coffee! will be around most of Thurs and early Sat. alternatively you can also come see me give a talk (at a museum apparently) on Fri:

cphnlp.github.io

Copenhagen NLP Symposium 2025

symposium website

cphnlp.github.io

7

Reposted by Najoung Kim

Anna Rogers @annarogers.bsky.social · May 26

📢 The Copenhagen NLP Symposium on June 20th!

- Invited talks by @loubnabnl.hf.co (HF) @mziizm.bsky.social (Cohere) @najoung.bsky.social (BU) @kylelo.bsky.social (AI2) Yohei Oseki (UTokyo)
- Exciting posters by other participants

Register to attend and/or present your poster at cphnlp.github.io /1

Copenhagen NLP Symposium 2025

symposium website

cphnlp.github.io

1 12 35

Reposted by Najoung Kim

Matt Goldrick @mattgoldrick.bsky.social · May 29

Very sorry to learn of the passing of LouAnn Gerken, who had such an impact on our understanding of the acquisition of speech sounds and sound patterns obits.mlive.com/us/obituarie...

LouAnn Gerken Obituary (1959 - 2025) - Tucson, AZ - Arizona Daily Star

View LouAnn A. Gerken's obituary, send flowers and sign the guestbook.

obits.mlive.com

2 26 48

Najoung Kim @najoung.bsky.social · May 4

thanks Matt, means a lot!! 😊

1

Najoung Kim @najoung.bsky.social · May 4

hello NAACL friends I'm giving a keynote today at RepL4NLP at 1:30PM local time, come say hi! I'll mostly be musing about things with light research discussions

Screenshot of a slide that says "what does it take to convince ourselves that a system is exhibiting compositionality?" with a side comment "mostly AI, but humans too!!" for the word system

1 1 18

Najoung Kim @najoung.bsky.social · Mar 29

🤫🤫🙈

2

Najoung Kim @najoung.bsky.social · Mar 29

we are scheming to make everyone's lab name a recursive acronym!

1 4

Najoung Kim @najoung.bsky.social · Mar 29

i think SNAIL is excellent

1 3

Najoung Kim @najoung.bsky.social · Mar 29

in liminal state, as correctly described by a colleague

2

Najoung Kim @najoung.bsky.social · Mar 27

so very excited that naomi is joining!!! a huge win for cds 💖

Naomi Saphra @nsaphra.bsky.social · Mar 27

Life update: I'm starting as faculty at Boston University
@bucds.bsky.social in 2026! BU has SCHEMES for LM interpretability & analysis, I couldn't be more pumped to join a burgeoning supergroup w/ @najoung.bsky.social @amuuueller.bsky.social. Looking for my first students, so apply and reach out!

CDS building which looks like a jenga tower

6

Najoung Kim @najoung.bsky.social · Feb 4

"Gaming Linguists" good bigram

Korean to English Gaming Linguists Urgently Required

1 4

Najoung Kim @najoung.bsky.social · Dec 18

Repost appreciated! 🙏

ACL 2025 Ling theory & Cognitive modeling track is looking for emergency reviewers. The emergency review period is between 3/18-26, and these reviewers will be excluded from the ARR cycle. If you're interested, please sign up here! docs.google.com/forms/d/1fH7...

ACL 2025 Ling theory & Cognitive modeling track emergency reviewer volunteer form

The Linguistic Theories, Cognitive Modeling, and Psycholinguistics track at ACL 2025 is looking for emergency reviewers. The emergency reviews will take place between 18th to 26th of March, 2025. Thes...

docs.google.com

11 4

Najoung Kim @najoung.bsky.social · Dec 1

have many tasks but immobilized by car

1 10

Reposted by Najoung Kim

Ben Lipkin @benlipkin.bsky.social · Nov 25

Lots of folks talking about scaling LLM inference over this last year

Internally, I’ve been developing and using a library that makes this extremely easy, and I decided to open-source it
Meet the decoding library: github.com/benlipkin/de...

1/7

GitHub - benlipkin/decoding: Composable inference algorithms with LLMs and programmable logic

Composable inference algorithms with LLMs and programmable logic - benlipkin/decoding

github.com

1 5 26

Najoung Kim @najoung.bsky.social · Nov 22

this is a first... suffering maximized...

The password must NOT contain a non alpha character in the last two positions of the password!

7

Najoung Kim @najoung.bsky.social · Nov 21

why say "bad" when you can say "ungood"

3