📖 "Explainability and Interpretability of Multilingual LLMs: A Survey" by @lucasresck2 et al. - The first comprehensive review of its kind! https://t.co/7s4k2lYrCW

🎉 #EMNLP2025 Exclusive Giveaway & More! 🎉 💥 500 prizes up for grabs — including AirPods 🎧, GPU credits, and more! Want in? 🚀 How to Enter: 1️⃣ Like and Retweet this post 2️⃣ Visit us at Booth https://t.co/KEX3mK9N3r to claim your spot in the lucky draw! 🔑 Limited spots -

Attending EMNLP 2025 in Suzhou, China this week? So is the CopeNLU group -- come find us at our paper presentations! ⤵️ https://t.co/3fdRaJjxSB @atanasovapepa @rnav_arora @whoSiddheshp, @_kire_kara_ @JingyiSun_ @Shoejoe_ @dustin_wright37 @lucasresck2 #NLProc #AI #EMNLP2025

A huge thank you to my supervisors for their invaluable feedback. As my first accepted PhD paper, this one means a lot! Looking forward to presenting in Suzhou, China 🇨🇳!

💡 Our key findings reveal: • Most work adapts existing, English-centric methods instead of creating novel ones for multilingual contexts • A significant skew against low-resource languages The paper also identifies key challenges and outlines promising avenues for research.

🗺️ In the survey, we explored the literature to categorise over 200 papers by: • Explainability techniques used • Multilingual tasks addressed • Languages investigated • Available resources

Thrilled to announce that my survey paper has been accepted at #EMNLP2025 Main! 🎉 To our knowledge, this is the first comprehensive survey dedicated to multilingual explainability. 📄 Preprint: https://t.co/vkHt8fzEjr With @annalkorhonen @IAugenstein #NLP #ExplainableAI

Amazing #LxMLS so far, with great talks by @mariotelfig @nlpnoah Slav Petrov @kchonyc! After the banquet and a free day to explore Lisbon, looking forward now to the next 3 days, with @danqi_chen @sarahookr Desmond Elliot @mohitban47 @adelehr @_albertgu.

With (most of) the #lxmls crowd.

We propose a novel contrastive-inspired loss function that incorporates human annotations into text classification. It significantly enhances the plausibility of model explanations, without causing substantial accuracy degradation. With @jpocom and @marcosmrai

Paper accepted to #NAACL2024 Findings! 💙 #NLP #XAI https://t.co/cSkVMMn5Y4

Assistant with Bard has appeared a bit early in the Pixel Tips app, giving us another look at how the upcoming feature will work. But it also reveals when it'll arrive and on what devices it'll be available on. This is early info and could be wrong, but here's what I found: 1)

Unfortunately, WhatsApp notifications on Wear OS doesn't respect phone silent notifications, including when phone screen is on. Impossible to continue using it until this is fixed

@WABetaInfo

The new WhatsApp for Wear OS is awesome: simple, useful, beautiful. Great update!

From what I understood of "Google passkeys," Google rebranded the existing Android security keys AND allowed password bypass on login. However I don't know the technicalities

O @nubank comprou a Olivia AI pra desativar esse aplicativo e integrar o próprio gerenciador de gastos no app do banco. Agora eles vão descontinuar esse serviço e avisam isso com quatro dias de antecedência. O cliente que se vire pra cuidar do seu dinheiro

Happy to share a new paper I co-authored, to be presented at @aistats_conf! DnX is a new GNN explainer which first distills the GNN and explains it afterwards. Experimental results are SOTA and we provide quite nice theoretical results. Thanks to all co-authors! #GNN #XAI

Só hoje percebi que a Elis Regina canta uma parte da música Como Nossos Pais, do Belchior, em terceira pessoa. Uma referência ao próprio Belchior?

Cientista demitido por Bolsonaro após questionar dados do desmatamento presidirá o CNPq https://t.co/T9dVhkBzTJ