Arnav Arora

Thrilled to have been awarded the Association for Computational Linguistics 2016 test of time award for my first ever paper, written with/under the guidance of @dirkhovy.bsky.social A couple of cute things about the paper/its genesis/its outcomes

Attention #NLProc researchers, the EACL 2027 website is officially LIVE: 2027.eacl.org! 🎉 🇬🇷 Join us in Athens, Greece (Mar 9-13, 2027) at #EACL2027 📅 ARR submission deadline: Aug 6, 2026. Open to all areas of CL/NLP + related fields. Stay tuned for the detailed CfP soon!

Big news! The 14th ACM Collective Intelligence Conference (CI 2026) will be held September 27-30, 2026, at the Virginia Tech Institute for Advanced Computing near Washington, DC. I'm excited to be a program co-chair this year. ci.acm.org/2026/ Papers and abstracts due 8 June. Check it out!

FYI #ACL2026 has an unusual registration system this year, and probably a lot of people who want to attend will not be able to. Spots are limited to 3.5k people, and only presenting authors can register during the first phase. Then, *if* there are spots left, others can try to register.

The paper acceptance notifications will be out by the 6th of April, AoE. The PCs are working hard throughout the holiday season to finalize the decisions. Apologies for the delay!

Back after a successful #EMNLP2025 conference in Suzhou, China -- some impressions ⤵️ Our papers: www.copenlu.com/news/8-paper... @apepa.bsky.social @rnv.bsky.social @siddesh.bsky.social @kirekara.bsky.social @shoejoe.bsky.social @zainmujahid.me @lucasresck.bsky.social @copenlu.bsky.social #NLProc

Turns out, a good way to reduce bias in LLMs is actually to make them more biased first. We came up with a neat way to do that using token based fine-tuning and then steering, getting some interesting results for both real world and fictional biases. Any feedback is welcome!

📢First Workshop on NLP4Democracy @ COLM 2025 📄Submit your non-archival abstracts by June 19 📅Attend the workshop in Montréal on Oct 10 for talks on: • Applying NLP to study democracy • Using LLM to improve democratic systems • AI-driven threats to democracy #NLP #Democracy #COLM2025 #AIforGood

Greta Thunberg, after trying to do the humanitarian work our governments are supposed to be doing: "Whatever the odds are, we have to keep trying…when you look at the state of the world, everything feels meaningless — but unless you try to do everything you can, that's when we lose our hope, no?"

Join us in Copenhagen for the Pre-ACL 2025 Workshop! 🇩🇰 We’re excited to welcome researchers and practitioners in Natural Language Processing, Generative AI, and Language Technology to a one-day workshop on 26 July 2025 – just ahead of ACL 2025 in Vienna. Learn more: www.aicentre.dk/events/pre-a...

This is a fantastic oral history of the last 10 years of NLP and AI. www.quantamagazine.org/when-chatgpt...

In "Investigating Human Values in Online Communities", we perform a high-scale study of the unique values expressed by online communities with different perspectives arxiv.org/abs/2402.14177 #NAACL2025 #NLProc @nadavb.bsky.social @rnv.bsky.social @frimelle.bsky.social @iaugenstein.bsky.social

C3NLP Workshop #NAACL2025: @iaugenstein.bsky.social will be presenting on tailoring LLM outputs to cultures, including where implicit cultural personalisation based on names leads to over-simplification arxiv.org/abs/2502.11995 @rnv.bsky.social @frimelle.bsky.social bsky.app/profile/frim... #NLProc

The CopeNLU group will be giving four paper presentations and one invited talk at #NAACL2025 this week, on topics including explainable AI and cross-cultural #NLProc Schedule + more details ⤵️ #XAI @apepa.bsky.social @rnv.bsky.social @nadavb.bsky.social @frimelle.bsky.social @iaugenstein.bsky.social

if you are using primarily local models, which ones are you using and for what lately?

Happy to share that I've joined the Apple Machine Learning Research team in Copenhagen as a research intern! Will continue to build on topics from my PhD, equitably advancing LLM access for all, working with @maartjeterhoeve.bsky.social and Natalie Schluter.

The high effort solution is to use an LLM to make a browser extension which tracks your academic reading and logs every paper you interact with to github, which builds and publishes a webapp to expose the data. Which, clearly only a crazy weirdo would do. dmarx.github.io/papers-feed/

This popped up on HN the other day, and it was one of the more fun “classical cryptography” posts I’ve seen in ages. Roughly speaking, someone discovered that AI models like Claude can decode the Caesar cipher, even when the “key” used is enormous. fi-le.net/byzantine/

Using simple, small models with the goal of usability and scalability of the task, we hope social scientists, journalists and researchers use this as a first step in studying multimodal framing and its intended/unintended effects. More here: bsky.app/profile/mari...

When we read the news, images can convey different things than text itself. Unlike other works which look at text, we study this as a “multimodal” framing problem & analyze where text and images communicate different “frames”. Checkout our paper here: arxiv.org/abs/2503.20960 @aicentre.dk

I'm so grateful to the British Computing Society & Bloomberg for honouring me with the Karen Spärck Jones Award 🙏 I gave the award lecture on LLMs’ Utilisation of Parametric & Contextual Knowledge at #ECIR2025 today (slides: isabelleaugenstein.github.io/slides/2025_...) www.bcs.org/membership-a...

New study with @iaugenstein.bsky.social’s group analyzing the interplay between photos and text in the news

New work on multimodal framing! 💫 Some fun results: comparisons of the same frame when expressed in images vs texts. When the "crime" frame is expressed in the article text, there are more political words in the text, but when the frame is expressed in the article image, more police words.

I have sort of given up on proactively updating the Danish Machine Learning Peoples starter pack. If anyone thinks that they should be added or know someone who should, please don't hesitate to comment here or sent a message 🤗. bsky.app/starter-pack...

I am still in need of emergency reviewers for ARR this cycle for the computational social science track, please DM me if you have capacity 🙏

NEW PAPER 📜 Shifting Perspectives: Steering Vector Ensembles for Robust Bias Mitigation in LLMs ArXiv: arxiv.org/abs/2503.05371 GitHub: github.com/groovychoons... Extremely Unofficial Blog Post: zarasiddique.com/blog/shiftin...

We had the pleasure of hosting @rnv.bsky.social for a talk on "NLP and CSS: lessons from one for the other". Thank you for your great talk!

📢New Paper Alert!🚀 Human alignment balances social expectations, economic incentives, and legal frameworks. What if LLM alignment worked the same way?🤔 Our latest work explores how social, economic, and contractual alignment can address incomplete contracts in LLM alignment🧵

New preprint! Metaphors shape how people understand politics, but measuring them (& their real-world effects) is hard. We develop a new method to measure metaphor & use it to study dehumanizing metaphor in 400K immigration tweets Link: bit.ly/4i3PGm3 #NLP #NLProc #polisky #polcom #compsocialsci 🐦🐦

When I say my name, people start speaking French to me, although my French is basic. That also happens with AI systems. We wrote a whole paper on that, testing across models for presumed cultural identity based on names w/ Siddhesh Pawar @rnv.bsky.social @iaugenstein.bsky.social

This report has a preview of the upcoming changes to the ARR cycle (not sure if these dates are confirmed or not)

Science is political. It wouldn't necessarily have to be partisan. But when one party is opposed to science-based policy, scientists have a choice between saying "you're right, thoughts and prayers are probably as effective as reducing carbon emissions" or saying "fuck that."

This is a very serious topic but making this are me feel better for about thirty seconds. 💙

#NLP question: is there a way to set email preferences for ARR/OpenReview? I am getting dozens of emails every day about interactions on papers I am SACing and would like to make my inbox more manageable…

Wow. it must feel terrible when someone takes something you made and uses it without permission or payment

When I grew up in Sweden in the 1990s the government distributed the book “Tell Ye Your Children”. Its purpose is to remind people how fascism, and seeing oneself as better than others can lead to inhuman evil. I think about it often and recommend you read it too.

Over the last few years, several people from the @milanlp.bsky.social lab went onto the academic job market in #NLP. Based on the conversations with them, I collected some observations we shared internally. In case they are useful for others, though, here they are: dirkhovy.com/post/2025_01...

Today, we are releasing MSTS, a new Multimodal Safety Test Suite for vision-language models! MSTS is exciting because it tests for safety risks *created by multimodality*. Each prompt consists of a text + image that *only in combination* reveal their full unsafe meaning. 🧵

Bit late to announce but I'm visting Dr. Kalina Bontcheva at Uni Sheffield for the next couple of months to work on disinformation related stuff! UK folks, do say 👋 if you're around or hosting for a talk. 😄

Counting down to the sanewashing by WaPo, LAT, ABC, Facebook, ...