Our pick of the week by Beomseok LEE @ INTERSPEECH 🇬🇷: "UniverSLU: Universal Spoken Language Understanding for Diverse Tasks with Natural Language Instructions" by Siddhant Arora, Hayato Futami, Yifan Peng, Roshan Sharma, Shinji Watanabe, et al., 2024. #SLU #languageunderstanding #speech

Last on our power panel: none other than Helena Moniz 🤩 President of EAMT and the International Association of Machine Translation. Currently Chair of the Ethics Committee of the Center for Responsible AI (centerforresponsible.ai) with Unbabel You don't want to miss this!

Our pick of the week by Dennis Fucci: "Encoding of lexical tone in self-supervised models of spoken language" by Gaofei Shen, Michaela Watkins afra alishahi, Arianna Bisazza at CogSci’24 🏙️, and Grzegorz Chrupała 🇪🇺🇺🇦, NAACL 2024. #Speech #SpokenLanguageModels #ToneEncoding #Interpretability #Phonology

The paper is finally available on arXiv! 🎊 ➡️ arxiv.org/abs/2405.08477

Less than a month until we see all you wonderful people at #GITT2024 EAMT2024 🤩 Time to send in your questions and topics for the panel that will wrap up the workshop. Share your ideas here: forms.gle/2kFXWYUGaPSmi6…

#EAMT2024 proceedings are out! Check them at eamt2024.sheffield.ac.uk/programme/proc… #NLProc #MachineTranslation

In Sheffield for #EAMT2024! Ready to present our research and engage with this amazing community Luisa Bentivogli BeatriceSavoldi Marco Gaido Andrea Piergentili EAMT2024

Andrea Piergentili hands down the best poster boaster so far 😂😂😂 #EAMT2024

Interested in how to enhance gender-inclusive #machinetranslation with LLMs? Come to #eamt2024 poster session at 13:45 and talk to our PhD student Andrea Piergentili! 📌Paper: arxiv.org/pdf/2405.08477 EAMT2024 BeatriceSavoldi Matteo Negri Luisa Bentivogli

We're big fans of poster boasters at #GITT2024 Janica Hackenbuchner Manuel Lardelli (he/him) Andrea Piergentili and Joke Daems presenting, come say hi at the posters later!

Very active poster discussions! #GITT2024

Aaand it's a wrap on #GITT2024 Thank you to everyone participating, great discussions and inspiration! See you at the third edition of GITT at some point in the future? 👀😉

'The Tail Wagging the Dog' (Nikil Selvam et al) is one of the most interesting papers I read recently. All of us working on AI ethics should be aware of the interaction of social and non-social biases in LMs. Looking forward to more research on it. aclanthology.org/2023.acl-short… MT Group at FBK

Our pick of the week by Andrea Piergentili: "The Tail Wagging the Dog: Dataset Construction Biases of Social Bias Benchmarks" by Nikil Selvam, Sunipa Dev, Daniel Khashabi 🕊️, tusharkhot, Kai-Wei Chang (ACL 2023). #bias #socialbias #benchmark #ACL2023NLP #NLP #NLProc

📢Come and join our group! We offer a fully funded 3-year PhD position with the IECS Doctorate School Università di Trento: Speech Translation in the LLM Era (Area A6) ⏱️Deadline: August 5th, 2024, h 4.00pm (CEST) 📌Application Details: iecs.unitn.it/education/admi… #NLProc Fondazione Bruno Kessler - FBK

Our pick of the week by BeatriceSavoldi: "Rethinking Model Evaluation as Narrowing the Socio-Technical Gap" by Vera Liao and Ziang Xiao, 2023. arxiv.org/abs/2306.03100 #Human #HumanCentered #Model #Evaluation #ModelEvaluation #AI

Meet the FBK MT group at #ACL2024 next week! 🐘 🥁Our Marco Gaido and Sara Papi will attend the conference and the IWSLT 2024 workshop and present the following papers! 📌More info is available on our website: mt.fbk.eu/our-postdocs-m… #ACL2024NLP #SPEECH #SQA #SFM #LLM #ST

Interesting take on gender bias in LLMs and retrievers: gender information is highly extractable but gender bias in RAG tasks is likely not due to the retrievers MultiContrievers: Analysis of Dense Retrieval Representations (Goldfarb-Tarrant et al., 2024) arxiv.org/abs/2402.15925

Our pick of the week by Andrea Piergentili: "MultiContrievers: Analysis of Dense Retrieval Representations" by Seraphina Goldfarb-Tarrant, Pedro Rodriguez, Jane Dwivedi-Yu, and Patrick Lewis, 2024. #NLProc #retrieval #genderbias #RAG #LLM

Today, at the DigitalIndustry-FBK Demo Day, Andrea Piergentili introduced our tool for Gender Neutral Translation with #LLMs. @fbk_research #NLProc