Looking for a new resource for studying misinformation in Italian? Check out IRMA!
Recently presented at #EACL2023 , IRMA is a corpus of over 600,000 Italian news articles from untrustworthy sources.

Generating intermediate steps improves symbolic reasoning accuracy – but why?

We investigate and factorize the benefit in our #EACL2023 Findings paper. Choice of reasoning strategies matters for performance! #NLProc

abs: arxiv.org/abs/2302.08148

So thrilled to have got the chance to present our work and my master’s thesis output (Exploring Segmentation Approaches for Neural Machine Translation of Code-Switched Egyptian Arabic-English Text ) during the poster session. #eacl 2023 #eacl aclanthology.org/2023.eacl-main…

Hmm, what do you think, is it #sigTyp or #LoResMT this morning? #EACL2023

Representing Rycolab at #eacl2023 with Butoi Alexandra Andreas Opedal Karolina Stanczak Clemente Pasti and Kevin Du!

workshops day: cats of #EACL2023 edition

🚨Join us on May 3rd to see our #eacl2023 poster and presentation on 'NusaX: Multilingual Parallel Sentiment Dataset for 10 Indonesian Local Languages'.

This is our initiative to create an NLP resource for underrepresented 🇮🇩Indonesian languages.

arxiv.org/abs/2205.15960

Isolated sign language recognition from video is hard! Why not help your model by showing it how signs can be broken down? Our paper 'Improving Sign Recognition with Phonology' shows that this idea can boost accuracy by nearly 10%! #EACL2023 1/3 arxiv.org/abs/2302.05759

Boom! 🤯

Our NusaX is awarded outstanding paper 😎
of #eacl2023 . We are happy that the NLP community recognizes our work on underrepresented languages 🇮🇩😀

It is a remarkable milestone for #indonlp community.

Thanks to all collaborators! #NLProc eaclmeeting

Dubrovnik might have the best sunset i’ve ever experienced 😭 #EACL2023

Tutorial on privacy-preserving NLP at #eacl2023 kicked-off by Sepideh Ghanavati showing why privacy is hard

Sticking around Dubrovnik for the EACL workshops?

Come check out our Findings paper on generating synthetic conversation datasets!

I'll be presenting a poster tomorrow (May 5) at the Insights from Negative Results in NLP workshop at 4:30pm, in Elafiti 4.

#EACL2023 #NLProc

Finished presenting MetaQA at #eacl 2023 #eacl thank you everybody for coming! Super honored for all the interest we got! 😊

Retrieval-augmented Language has shown great performance on knowledge-intensive tasks. In our recent #EACL2023 paper QA-Memory (arxiv.org/abs/2204.04581), we explore the possibility to augment pre-trained language models (T5) with QA-pairs as memory. QA pairs as the knowledge ..

Walking the Walls of Dubrovnik after the end of the main conference of #EACL2023 !

Excited for workshops tomorrow!!

Excited for EACL 2023 eaclmeeting. Here is a summary of our posters. See you at the poster session if you are attending!
#eacl2023 #NLP roc #NLP #LLM

Such a shame that both of my posters at #EACL2023 were at the same time, and I couldn't talk to everyone who came up to my stands! But I hope ye enjoyed the designs at least 🫏👑

I was asked multiple times how I made them, so here is a thread on poster design Academic Chatter™

Thrilled to be at #EACL2023 this week! Join me on Wednesday (9am local time) as I present our research (w/ Lena Voita and Andre Martins) on hallucinations in neural machine translation!

This is how I presented my paper at the #eacl2023 / #NLProc conference today during the air-ride alert, sitting on the floor, joined by my fellow Mrs. Bath Towel and Mr. Toilet Duck.

Now happening: our emotion analysis tutorial at #EACL2023 with Sanja Stajner Language Tech