Talk to rachneet about our work eaclmeeting!

We are proud to announce that the contribution »Sensitivity, Performance, Robustness: Deconstructing the Effect of Sociodemographic Prompting« by Tilman Beck, Hendrik Schuff, Anne Lauscher (she/her) (Universität Hamburg) and Iryna Gurevych (UKP Lab) has just been awarded the #EACL2024 Social Impact Award!

Five papers authored or co-authored by UKP members have been accepted for publication at this year's @NAACLmeeting, the Conference of the North American Chapter of the ACL (@NAACL). Congratulations to everyone involved! 💐 #NAACL2024 informatik.tu-darmstadt.de/ukp/ukp_home/u…

So happy to meet so many amazing people at #EACL2024 in 🇲🇹! We presented our framework for differentially private machine translation with MSc student Doan Nam Long Vu, find out more about our paper "DP-NMT: Scalable Differentially-Private Machine Translation" below ⬇️

Good at language 🟰 good at thinking? Not true for LMs! Even if we might think that when talking to them 💬 Meet Holmes 🔎, a benchmark to assess linguistic competence📚 of LMs untwined from other skills 🔥 Spoiler 🤫 Architecture is one key! (1/🧵) 🌐 holmes-benchmark.github.io

Follow the minds behind this research Andreas Waldis, Yotam Perlitz 👾, Leshem (Legend) Choshen 🤖🤗, Yufang Hou and Iryna Gurevych (UKP Lab, @HSLU, IBM Research, MIT CSAIL, MIT-IBM Watson AI Lab) for further insights and discussion. (9/9) 📄: arxiv.org/abs/2404.18923 🌐: holmes-benchmark.github.io

We are happy to announce that 11 papers authored or co-authored by UKP members have been accepted for publication at this year's ACL 2025 in Bangkok 🇹🇭. Congratulations to everyone involved! 💐 #ACL2024NLP 📑 informatik.tu-darmstadt.de/ukp/ukp_home/u… The accepted papers are (1/🧵):

Happening this upcoming Monday at UKP Lab

@CarinaKauf’s new work overturns our earlier claim (tinyurl.com/bdzxpfvp) that local semantic composition drives the language system. New data show that linguistic inputs must be syntactically parsable to fully engage the lang system (and they need not be semant. plausible).

The story of real-world Fact-Checking at UKP Lab 🕵️ Current NLP fact-checking approaches are unrealistic due to missing counter-evidence: 🔎 aclanthology.org/2022.emnlp-mai… NLP fact-checking should model ambiguous real-world claims: 🔎 aclanthology.org/2024.tacl-1.1/ – 🧵 (1/6)

Please drop by our poster tomorrow and talk to my co-authors about how IVON can directly optimize a variational objective with similar performance and cost as Adam but many things on top (uncertainty, model adaptation, diagnosis, ...) Thu. 13:30-15:00, hall c 4-9, poster #1402

Announcing the ICLR 2025 Call for Papers! Abstract submission: 11:59pm, Sept 27 (AoE) Submission date: 11:59pm, Oct 1 (AoE) Reviews released: Nov 12 Author/Reviewer Discussion: Nov 12-26 Author Last Day to Reply: Nov 27 Final Decisions: Jan 22 2025 iclr.cc/Conferences/20…

Tonight, I'm flying out to attend #ACL2024NLP! Come to Poster Session 4 on August 13 at 10:30AM to discuss our paper 💥Mission: Impossible Language Models💥 I'm also happy to just make a new friend or be your personal photographer at your poster session! See you in Bangkok 🇹🇭

If you're interested in emergent abilities and LLMs, come stop by our poster!

Absolutely wonderful to be working with the incredible UKP Lab led by Iryna Gurevych 🎉 Looking forward to several more amazing discussions and collaborations ahead 🙌 #NLProc #ACL2024NLP

A video summary of our ACL 2025 #ACL2024NLP work demonstrating that 🤖🤖Emergent Abilities in LLMs are just In-Context Learning Collaborative work between BathNLP and UKP Lab's Sheng Lu, Irina Bigoulaeva , rachneet , and Iryna Gurevych ! More at: x.com/harish/status/… Read

We won OUTSTANDING PAPER at #ACL2024 for our work on evaluating values and opinions in LLMs 🥳 Thank you to the reviewers and awards committee, and again to my amazing co-authors, especially joint first author Valentin Hofmann Check out the poster + paper + full author list below 👇

One of our highlights at this year's @ACLmeeting: The presentation of our paper on Emergent Abilities in LLMs, authored by Sheng Lu, Irina Bigoulaeva, rachneet, @Harish Tayyar Madabushi and Iryna Gurevych (BathNLP | UKP Lab). #ACL2024NLP x.com/UKPLab/status/…

Our colleague in action: At #ACL2024NLP in 🇹🇭 Furkan Şahinuç presented a paper on Systematic Task Exploration with LLMs, co-authored with Ilia Kuznetsov, @YufangHou and Iryna Gurevych (UKP Lab | IBM Research Ireland | InterText@UKP Lab | hessian.AI). x.com/UKPLab/status/…

This is a work fantastically led by Haritz Puerto . Please reach out for more questions. I will be attending #EMNLP2024 as well. Please reach out to discuss 1) #codeprompting, 2) ERVQA (Large VLMs in healthcare), 3) Affordance knowledge and "grounding" ability of #LLMs and #VLMs