This is an exciting new workshop at CHI that I will be at! Deadline for submissions is one week away.
06.02.2026 00:04 β π 1 π 1 π¬ 0 π 0@jkkummerfeld.bsky.social
NLP faculty - University of Sydney he/him (this account is for professional topics only) https://www.jkk.name
This is an exciting new workshop at CHI that I will be at! Deadline for submissions is one week away.
06.02.2026 00:04 β π 1 π 1 π¬ 0 π 0Do you have ideas for the future of reading?
Submit a 2-4 page paper to the CHI workshop I am co-organising! (deadline Feb 12) βScience and Technology for Augmenting Reading"
chi-star-workshop.github.io
π¨ @jkkummerfeld.bsky.social from our school is organising "The 23rd Annual Workshop of the Australasian Language Technology Association" to be held between 26/11/2025 - 28/11/2025 at #USyd @sydney.edu.au
Open to Students, academics, and industry working on NLP
To register: alta2025.alta.asn.au
π’ We are hiring! The School of #ComputerScience of #USyd π¦πΊ is seeking outstanding women academics, at all seniority levels, for continuing (equivalent tenure system) positions.
Deadline to apply: β° August 25 #AcademicJobs #WomenInCS
usyd.wd105.myworkdayjobs.com/en-GB/USYD_E...
UG or PG student at @sydney.edu.au or another Australian university? Come do a 6-week paid Vacation #Research #Internship (VRI) with our academics this Winter, choosing among the 46 projects proposed at the School of #ComputerScience! β° Deadline: April 21.
Details: www.sydney.edu.au/engineering/...
The ARR site is now updated and the committee report is here: www.aclweb.org/adminwiki/im...
21.02.2025 00:14 β π 2 π 0 π¬ 0 π 0A few thoughts on why DeepSeek's ideas are likely to lead to more use of AI (and hence demand for chips), not less
30.01.2025 23:21 β π 4 π 0 π¬ 1 π 0I found the episode very interesting, and am curious whether you would be open to having someone on from a CS/ML background to give a perspective on the same question? (e.g., an NLP academic based in CS) Of course, I realise the podcast is *ling*thusiasm, so that may not fit with your approach :)
26.11.2024 23:29 β π 1 π 0 π¬ 0 π 0With the debate over peer review going strong, what other classic topics should be revived here?
vim vs. emacs?
spaces vs. tabs?
Star Trek vs. Star Wars?
The Oxford comma?
The jokeβs on both of you - the post is already available on an open access server!
25.11.2024 11:18 β π 3 π 0 π¬ 0 π 0More broadly, part of the philosophy of ARR is that we will change / improve, including via community feedback (lots of changes have happened since it started).
We are just one corner of AI though :)
In the case of this survey, it was advertised to all ACL members, all authors and reviewers in ARR, and on social media, so hopefully we got a representative sample. I havenβt seen the results yet, but am very curious!
25.11.2024 08:17 β π 5 π 0 π¬ 1 π 0If you want to help improve peer review, we are looking for a new Co-CTO for ACL Rolling Review!
Requirements:
- Post-PhD
- Experienced with Python (including command line use)
- Time commitment of 3 hours a week on average (but note that you are not expected to review while serving)
Contact me!
You forgot to remind people - we are hiring!
20.11.2024 09:16 β π 1 π 0 π¬ 1 π 0Maybe we are doing too good a job of teaching computational thinking and so their minds have become entirely discrete
17.11.2024 21:27 β π 1 π 0 π¬ 0 π 0Integration by parts is definitely in the Australian high school curriculum, but maybe only for extension mathematics? Should be covered in first year calculus though
17.11.2024 21:04 β π 1 π 0 π¬ 1 π 0(And now I will return to my usual NLP postingβ¦ :D)
11.11.2024 08:16 β π 1 π 0 π¬ 0 π 0If you want an iPhone only lightweight time tracking app, I wrote this one: apps.apple.com/au/app/fast-...
The key missing feature for me at the time in other apps was the ability to change the timer without unlocking the screen. So I made an app with a widget to let me do that.
Another approach would be to give people a list of words and ask them to rate them as plausible continuations, but that seems hard to do at a statistically meaningful scale
09.11.2024 21:28 β π 4 π 0 π¬ 1 π 0True, as people will go for the most likely thing, and virtually no one will choose an option that we all agree is plausible but unlikely.
09.11.2024 21:27 β π 1 π 0 π¬ 1 π 0I also do this as a game in my first lecture of semester - I say a phrase and the whole class has to shout out the next word
09.11.2024 21:14 β π 1 π 0 π¬ 1 π 0I would expect a linguistics paper somewhere that has done this (possibly with a restricted vocabulary)
09.11.2024 21:14 β π 2 π 0 π¬ 1 π 0Sydney is a great place for research, teaching, and life! Join us :)
09.11.2024 08:08 β π 7 π 4 π¬ 0 π 0ICYMI: The School of Computer Science at the University of Sydney π¦πΊ is hiring, with multiple (equiv.) tenure-track positions. All areas of CS welcome, with a specific focus on Systems, PL, ML, and Quantum Computing.
usyd.wd3.myworkdayjobs.com/en-US/USYD_E...
Please share, and get in touch!
LaTeX: bringing people and software packages together since 1984
09.11.2024 08:05 β π 2 π 0 π¬ 0 π 0Also curious about this!
07.11.2024 09:05 β π 0 π 0 π¬ 1 π 0On my way to CHI! Excited about both papers I contributed to (links go elsewhere, sorry!):
An AI-resilient summarisation method
x.com/ziweigu/stat...
Finding ways to look at 10s-100s of LLM outputs at once
x.com/katyilonka/s...
If you will be there and want to meet up, let me know!
EMNLP 2024 - Miami! (Nov 12-16)
10.12.2023 08:46 β π 1 π 0 π¬ 0 π 0Fascinating keynote on fallacies in benchmarking by Adina Williams (ai.meta.com/people/adina...). Short version: in our excitement about LLMs we shouldnβt forget scientific principles of evaluation (Iβm definitely going to check out the stack of EMNLP papers Dr Williams is an author of!)
06.12.2023 04:08 β π 3 π 0 π¬ 0 π 0Computer Science at Sydney is hiring! (deadline Nov 30)
We get excellent students, have fantastic quality of life, and you get to be my colleague (hopefully that's a positive weight feature...).
If you aren't on the job market, please think about if you know someone who is and share the ad :)