#acl2025 — Public Fediverse posts

Tại hội nghị ACL 2025 ở Vienna (Áo) – sự kiện hàng đầu thế giới về xử lý ngôn ngữ tự nhiên, một nghiên cứu về bộ công cụ đánh giá mô hình ngôn ngữ tiếng Việt do nhóm kỹ sư Zalo phát triển đã thu hút sự chú ý của nhiều nhà khoa học quốc tế. Đại diện nhóm là nữ kỹ sư Việt Nam, góp phần mang công nghệ nội địa vươn ra thế giới. Đóng góp này khẳng định vị thế ngày càng cao của Việt Nam trong lĩnh vực AI và xử lý ngôn ngữ.

#AI #XuLyNguonNguTuNhien #Zalo #CongNgheViet #ACL2025 #WomenInTech #NghienCu

And consider following the authors Hiba Arnaout (UKP Lab /TU Darmstadt), Noy Sternlicht (HUJI), Tom Hope (HUJI & Allen Institute for AI), and Iryna Gurevych (UKP Lab/TU Darmstadt) if you are interested in more information or an exchange of ideas.

#NLProc #LLMs #SciNLP #ACL2025 #CitationAnalysis

And consider following the authors Hiba Arnaout (UKP Lab /TU Darmstadt), Noy Sternlicht (HUJI), Tom Hope (HUJI & Allen Institute for AI), and Iryna Gurevych (UKP Lab/TU Darmstadt) if you are interested in more information or an exchange of ideas.

#NLProc #LLMs #SciNLP #ACL2025 #CitationAnalysis

Many thanks to our colleagues for their contributions and to the organizers for a great conference — we’re already looking forward to the next #ACL!

📸 First photo by Andreas Geiger – thank you!

#ACL2025 #NLProc

Many thanks to our colleagues for their contributions and to the organizers for a great conference — we’re already looking forward to the next #ACL!

📸 First photo by Andreas Geiger – thank you!

#ACL2025 #NLProc

𝗧𝗵𝗿𝗼𝘄𝗯𝗮𝗰𝗸 𝘁𝗼 #ACL2025 𝗶𝗻 𝗯𝗲𝗮𝘂𝘁𝗶𝗳𝘂𝗹 𝗩𝗶𝗲𝗻𝗻𝗮 🍰✨

The UKP Lab had a fantastic time at this year’s #ACL. Our team presented a total of 𝟭𝟬 𝗽𝗮𝗽𝗲𝗿𝘀 in the Main and Findings tracks and 𝟮 𝗧𝗔𝗖𝗟 journal papers, sharing our latest research with the global #NLP community.

𝗧𝗵𝗿𝗼𝘄𝗯𝗮𝗰𝗸 𝘁𝗼 #ACL2025 𝗶𝗻 𝗯𝗲𝗮𝘂𝘁𝗶𝗳𝘂𝗹 𝗩𝗶𝗲𝗻𝗻𝗮 🍰✨

The UKP Lab had a fantastic time at this year’s #ACL. Our team presented a total of 𝟭𝟬 𝗽𝗮𝗽𝗲𝗿𝘀 in the Main and Findings tracks and 𝟮 𝗧𝗔𝗖𝗟 journal papers, sharing our latest research with the global #NLP community.

Retour de conférence à Vienne #ACL2025 en Flixbus pour cause de pas vraiment de possibilité de train...
PLUS JAMAIS CA !! C'était juste l'enfer (sale, bruyant, pas confortable) et on a été arrêtés DEUX FOIS par la douane (dont un contrôle total, y compris des bagages, 45 min de retard, merci)
.
A noter quand même, les chauffeureuses dudit bus, tou·te·s roumain·e·s et super débrouilles et sympas (bon, les gueulantes à 3h du mat, on s'en passerait, mais bon).

Ce n'est plus possible de pisser sous la douche (ie faire de son mieux pour ne pas prendre l'avion), il faut une politique des transports qui tienne debout.

Yuqicheng Zhu (@yuqichengzhu.bsky.social) presented our paper, Predicate-Conditional Conformalized Answer Sets for Knowledge Graph Embeddings, at #ACL2025. This paper studies a key question for the reliability of applications using knowledge graph embeddings: How to quantify the uncertainty? We show that not all data have the same uncertainty and propose a method to have stronger guarantees by grouping similar data.

https://arxiv.org/abs/2505.16877

#KnowledgeGraphs #MachineLearning

Yuqicheng Zhu (@yuqichengzhu.bsky.social) presented our paper, Predicate-Conditional Conformalized Answer Sets for Knowledge Graph Embeddings, at #ACL2025. This paper studies a key question for the reliability of applications using knowledge graph embeddings: How to quantify the uncertainty? We show that not all data have the same uncertainty and propose a method to have stronger guarantees by grouping similar data.

https://arxiv.org/abs/2505.16877

#KnowledgeGraphs #MachineLearning

If you share #NLProc research interests with me, my conference report for @aclmeeting.bsky.social #ACL2025 might be interesting for you – I highlight the papers that I personally found particularly relevant for our own research. www.romanklinger.de/blog/2025-08...

🤖 Do LLMs reflect our moral expressions—or alter them?

In our #ACL2025 paper, we study 12 widely used LLMs and found that moral expressions are altered to varying degrees depending on the editing task and moral conditioning prompt.

📍 ACL2025 — Hall X5 (28)
🕚 Today, 11:00–12:30

🤖 Do LLMs reflect our moral expressions—or alter them?

In our #ACL2025 paper, we study 12 widely used LLMs and found that moral expressions are altered to varying degrees depending on the editing task and moral conditioning prompt.

📍 ACL2025 — Hall X5 (28)
🕚 Today, 11:00–12:30

Wow, #acl2025 is already a super hot tag for this week's conf in Vienna (must read if you are an NLP person)

Team ISE (Mary Ann Tan and Shufan Jiang) attending #ACL2025 in Vienna to present our paper at the Fifth Workshop on Scholarly Document Processing at ACL 2025

Shufan Jiang, Mary Ann Tan, Harald Sack: MathD2: Towards Disambiguation of Mathematical Terms: https://aclanthology.org/2025.sdp-1.3/

@shufan @fiz_karlsruhe #NLP #AI @lysander07

Team ISE (Mary Ann Tan and Shufan Jiang) attending #ACL2025 in Vienna to present our paper at the Fifth Workshop on Scholarly Document Processing at ACL 2025

Shufan Jiang, Mary Ann Tan, Harald Sack: MathD2: Towards Disambiguation of Mathematical Terms: https://aclanthology.org/2025.sdp-1.3/

@shufan @fiz_karlsruhe #NLP #AI @lysander07

This time it's a home run – find me at ACL in Vienna!

https://2025.aclweb.org/

#acl #acl2025

On my way to Vienna for #ACL2025. Hit me up if you want to talk about Legal NLP, Perspectivism and Disagreement, AI for Social Good or other stuff.

On my way to Vienna for #ACL2025. Hit me up if you want to talk about Legal NLP, Perspectivism and Disagreement, AI for Social Good or other stuff.

Next week, I will be attending #ACL2025 to present a paper on generating prosodically-accurate Latin poetry recordings with LLMs. Pretty excited to be there for the first time! Check it out: https://aclanthology.org/2025.acl-srw.48/

Next week, I will be attending #ACL2025 to present a paper on generating prosodically-accurate Latin poetry recordings with LLMs. Pretty excited to be there for the first time! Check it out: https://aclanthology.org/2025.acl-srw.48/

Work by: Max Glockner (UKP Lab), Xiang Jiang (Amazon AGI), Leonardo F. R. Ribeiro (Amazon AGI), Iryna Gurevych (UKP Lab), and Markus Dreyer (Amazon AGI).

📄 Paper: https://arxiv.org/abs/2505.05949
💾 Data: https://huggingface.co/datasets/mglockner/neoqa
💻 Code: https://github.com/amazon-science/neoqa

#ACL2025 #NLProc #EvidenceBasedAI #LLM

(3/3)

The original post was published on Twitter/X by Markus Dreyer:
https://x.com/markusdr/status/1924873660969652306

Work by: Max Glockner (UKP Lab), Xiang Jiang (Amazon AGI), Leonardo F. R. Ribeiro (Amazon AGI), Iryna Gurevych (UKP Lab), and Markus Dreyer (Amazon AGI).

📄 Paper: https://arxiv.org/abs/2505.05949
💾 Data: https://huggingface.co/datasets/mglockner/neoqa
💻 Code: https://github.com/amazon-science/neoqa

#ACL2025 #NLProc #EvidenceBasedAI #LLM

(3/3)

The original post was published on Twitter/X by Markus Dreyer:
https://x.com/markusdr/status/1924873660969652306

Also consider following the authors Tianyu Yang (Ubiquitous Knowledge Processing (UKP) Lab, hessian.AI)‬, Xiaodan Zhu (Department of Electrical and Computer Engineering & Ingenuity Labs Research Institute, Queen's University), and Iryna Gurevych (Ubiquitous Knowledge Processing (UKP) Lab).

(5/5)

#NLProc #ACL2025 #TextAnonymization #LLMSafety #AIPrivacy

Also consider following the authors Tianyu Yang (Ubiquitous Knowledge Processing (UKP) Lab, hessian.AI)‬, Xiaodan Zhu (Department of Electrical and Computer Engineering & Ingenuity Labs Research Institute, Queen's University), and Iryna Gurevych (Ubiquitous Knowledge Processing (UKP) Lab).

(5/5)

#NLProc #ACL2025 #TextAnonymization #LLMSafety #AIPrivacy

Also consider following the authors Aniket Pramanick (Ubiquitous Knowledge Processing (UKP) Lab)‬, Yufang Hou (IT:U- Interdisciplinary Transformation University Austria, IBM Research), Saif M Mohammad (National Research Council Canada / Conseil national de recherches Canada), and Iryna Gurevych (Ubiquitous Knowledge Processing (UKP) Lab).

🗺️ See you at #ACL2025 in Vienna

(2/2)

#NLProc #ACL2025 #AI4Science #ACL2025

Also consider following the authors Aniket Pramanick (Ubiquitous Knowledge Processing (UKP) Lab)‬, Yufang Hou (IT:U- Interdisciplinary Transformation University Austria, IBM Research), Saif M Mohammad (National Research Council Canada / Conseil national de recherches Canada), and Iryna Gurevych (Ubiquitous Knowledge Processing (UKP) Lab).

🗺️ See you at #ACL2025 in Vienna

(2/2)

#NLProc #ACL2025 #AI4Science #ACL2025

Also consider following the authors Aniket Pramanick (Ubiquitous Knowledge Processing (UKP) Lab)‬, Yufang Hou (IT:U- Interdisciplinary Transformation University Austria, IBM Research), Saif M Mohammad (National Research Council Canada / Conseil national de recherches Canada), and Iryna Gurevych (Ubiquitous Knowledge Processing (UKP) Lab).

🗺️ See you at #ACL2025 in Vienna

(2/2)

#NLProc #ACL2025 #AI4Science #ACL2025

Also consider following the authors Aniket Pramanick (Ubiquitous Knowledge Processing (UKP) Lab)‬, Yufang Hou (IT:U- Interdisciplinary Transformation University Austria, IBM Research), Saif M Mohammad (National Research Council Canada / Conseil national de recherches Canada), and Iryna Gurevych (Ubiquitous Knowledge Processing (UKP) Lab).

🗺️ See you at #ACL2025 in Vienna

(2/2)

#NLProc #ACL2025 #AI4Science #ACL2025

Also consider following the authors Aniket Pramanick (Ubiquitous Knowledge Processing (UKP) Lab)‬, Yufang Hou (IT:U- Interdisciplinary Transformation University Austria, IBM Research), Saif M Mohammad (National Research Council Canada / Conseil national de recherches Canada), and Iryna Gurevych (Ubiquitous Knowledge Processing (UKP) Lab).

🗺️ See you at #ACL2025 in Vienna

(2/2)

#NLProc #ACL2025 #AI4Science #ACL2025

And consider following the authors Yongxin Huang (Ubiquitous Knowledge Processing (UKP) Lab), Kexin Wang (UKP Lab), Goran Glavaš (Julius-Maximilians-Universität Würzburg) and Iryna Gurevych (UKP Lab) if you are interested in more information or an exchange of ideas.

(3/3)

See you in Vienna! #ACL2025

And consider following the authors Yongxin Huang (Ubiquitous Knowledge Processing (UKP) Lab), Kexin Wang (UKP Lab), Goran Glavaš (Julius-Maximilians-Universität Würzburg) and Iryna Gurevych (UKP Lab) if you are interested in more information or an exchange of ideas.

(3/3)

See you in Vienna! #ACL2025

🧑‍🔬 Ever wondered what goes through a peer reviewer's mind? Can LLMs actually think like academic reviewers? 𝗦𝗧𝗥𝗜𝗖𝗧𝗔 (Structured Reasoning in Critical Text Assessment) has the answer 🧵

🔗Project: https://ukplab.github.io/acl2025-stricta/
📄 Paper: https://arxiv.org/pdf/2409.05367
💻 Code: https://github.com/UKPLab/acl2025-stricta

#NLP4PeerReview #HumanCentered #NLProc #ACL2025

(1/🧵 )

🧑‍🔬 Ever wondered what goes through a peer reviewer's mind? Can LLMs actually think like academic reviewers? 𝗦𝗧𝗥𝗜𝗖𝗧𝗔 (Structured Reasoning in Critical Text Assessment) has the answer 🧵

🔗Project: https://ukplab.github.io/acl2025-stricta/
📄 Paper: https://arxiv.org/pdf/2409.05367
💻 Code: https://github.com/UKPLab/acl2025-stricta

#NLP4PeerReview #HumanCentered #NLProc #ACL2025

(1/🧵 )

‪
It is official, our two long papers at #ACL2025 have now been published. Common work with Arne Rubehn (Concept Embeddings), and Frederic Blum and
Steffen Herbold
(Automated Language Affiliation).

https://aclanthology.org/2025.acl-long.1004/

https://aclanthology.org/2025.acl-long.876/

And consider following the authors Jiahui Geng (MBZUAI), Thy Thy Tran (UKP Lab/Technische Universität Darmstadt), Preslav Nakov (MBZUAI), and Iryna Gurevych (UKP Lab & MBZUAI)

See you in Vienna! #ACL2025 !

(4/4)

#MLLM #AISafety #Jailbreak #Multimodal #ConInstruction #ACL2025 #LLMRedTeaming #VisionLanguage #AudioLanguage#NLProc

And consider following the authors Jiahui Geng (MBZUAI), Thy Thy Tran (UKP Lab/Technische Universität Darmstadt), Preslav Nakov (MBZUAI), and Iryna Gurevych (UKP Lab & MBZUAI)

See you in Vienna! #ACL2025 !

(4/4)

#MLLM #AISafety #Jailbreak #Multimodal #ConInstruction #ACL2025 #LLMRedTeaming #VisionLanguage #AudioLanguage#NLProc

And consider following the authors Jiahui Geng (MBZUAI), Thy Thy Tran (UKP Lab/Technische Universität Darmstadt), Preslav Nakov (MBZUAI), and Iryna Gurevych (UKP Lab & MBZUAI)

See you in Vienna! #ACL2025 !

(4/4)

#MLLM #AISafety #Jailbreak #Multimodal #ConInstruction #ACL2025 #LLMRedTeaming #VisionLanguage #AudioLanguage#NLProc

And consider following the authors Jiahui Geng (MBZUAI), Thy Thy Tran (UKP Lab/Technische Universität Darmstadt), Preslav Nakov (MBZUAI), and Iryna Gurevych (UKP Lab & MBZUAI)

See you in Vienna! #ACL2025 !

(4/4)

#MLLM #AISafety #Jailbreak #Multimodal #ConInstruction #ACL2025 #LLMRedTeaming #VisionLanguage #AudioLanguage#NLProc

And consider following the authors Jiahui Geng (MBZUAI), Thy Thy Tran (UKP Lab/Technische Universität Darmstadt), Preslav Nakov (MBZUAI), and Iryna Gurevych (UKP Lab & MBZUAI)

See you in Vienna! #ACL2025 !

(4/4)

#MLLM #AISafety #Jailbreak #Multimodal #ConInstruction #ACL2025 #LLMRedTeaming #VisionLanguage #AudioLanguage#NLProc

And consider following the authors Sukannya Purkayastha (Ubiquitous Knowledge Processing (UKP) Lab), ‪Zhuang Li (RMIT University)‬, ‪Anne Lauscher‬ (Universität Hamburg), Lizhen Qu (Department of Data Science and Artificial Intelligence - Monash Information Technology), Iryna Gurevych (Ubiquitous Knowledge Processing (UKP) Lab) if you are interested in more information or an exchange of ideas.

(2/2)

See you in Vienna! #ACL2025

And consider following the authors Sukannya Purkayastha (Ubiquitous Knowledge Processing (UKP) Lab), ‪Zhuang Li (RMIT University)‬, ‪Anne Lauscher‬ (Universität Hamburg), Lizhen Qu (Department of Data Science and Artificial Intelligence - Monash Information Technology), Iryna Gurevych (Ubiquitous Knowledge Processing (UKP) Lab) if you are interested in more information or an exchange of ideas.

(2/2)

See you in Vienna! #ACL2025

🚨 𝗟𝗮𝘇𝘆 𝗧𝗵𝗶𝗻𝗸𝗶𝗻𝗴 𝗶𝘀 𝘁𝗵𝗲 𝗰𝗼𝗿𝗲 𝗿𝗲𝗮𝘀𝗼𝗻 𝗳𝗼𝗿 𝗯𝗮𝗱 𝗿𝗲𝘃𝗶𝗲𝘄 𝗾𝘂𝗮𝗹𝗶𝘁𝘆! 💤
At #ACL 2023, 24.3% of author-reported issues cited heuristic-based reviews 🤯. To solve this issue, we release #LazyReview, a new dataset to expose and understand this trend along with the models! 🔍📉

🖥️ Project: https://ukplab.github.io/acl2025-lazy-review/
📄 Paper: https://arxiv.org/pdf/2504.11042
💻 Code: https://github.com/UKPLab/acl2025-lazy-review

(1/🧵)

#PeerReview #NLProc #ACL2025

🚨 𝗟𝗮𝘇𝘆 𝗧𝗵𝗶𝗻𝗸𝗶𝗻𝗴 𝗶𝘀 𝘁𝗵𝗲 𝗰𝗼𝗿𝗲 𝗿𝗲𝗮𝘀𝗼𝗻 𝗳𝗼𝗿 𝗯𝗮𝗱 𝗿𝗲𝘃𝗶𝗲𝘄 𝗾𝘂𝗮𝗹𝗶𝘁𝘆! 💤
At #ACL 2023, 24.3% of author-reported issues cited heuristic-based reviews 🤯. To solve this issue, we release #LazyReview, a new dataset to expose and understand this trend along with the models! 🔍📉

🖥️ Project: https://ukplab.github.io/acl2025-lazy-review/
📄 Paper: https://arxiv.org/pdf/2504.11042
💻 Code: https://github.com/UKPLab/acl2025-lazy-review

(1/🧵)

#PeerReview #NLProc #ACL2025

And consider following the authors Chen Cecilia Liu, Anna Korhonen, and Iryna Gurevych if you are interested in more information or an exchange of ideas.

See you in Vienna! #ACL2025

#ResponsibleAI #CulturalNLP #NLProc #ACL2025

And consider following the authors Chen Cecilia Liu, Anna Korhonen, and Iryna Gurevych if you are interested in more information or an exchange of ideas.

See you in Vienna! #ACL2025

#ResponsibleAI #CulturalNLP #NLProc #ACL2025

💃 #WebDancer uses #ReAct framework for autonomous information seeking with 64.1% Pass@3 on GAIA
🚶 #WebWalker benchmarks web traversal capabilities, accepted at #ACL2025 conference
🧠 Features 4-stage training: browsing data construction, trajectory sampling, SFT & #RL

💃 #WebDancer uses #ReAct framework for autonomous information seeking with 64.1% Pass@3 on GAIA
🚶 #WebWalker benchmarks web traversal capabilities, accepted at #ACL2025 conference
🧠 Features 4-stage training: browsing data construction, trajectory sampling, SFT & #RL

🎉 The countdown to #ACL2025 in Vienna is on - and we’re excited to contribute 10 accepted papers and two #TACL articles from the UKP Lab!

In the days leading up to the conference, we’ll be sharing each of our contributions, including links to the arXiv preprints, code repositories, and more.
(1/🧵)

🎉 The countdown to #ACL2025 in Vienna is on - and we’re excited to contribute 10 accepted papers and two #TACL articles from the UKP Lab!

In the days leading up to the conference, we’ll be sharing each of our contributions, including links to the arXiv preprints, code repositories, and more.
(1/🧵)