Can LLMs encode biased expectations about men?
In our ACL Findings 2026 paper, “Real Men are Tough”, we study whether LLMs rely on traditional masculinity norms as expectations about male agents.
#NLProc #ACL2026
@ElisaLeonardell @sara_hlt @camillacasula @BoglarkaNyul
Across models, we find a dissociation: they rarely endorse traditional masculinity norms when asked directly, but often attribute MRNI-aligned behaviors to men in scenario-based inference, especially under ambiguity.
We release MRNI-BB to support further work on male-directed gender bias and on how gender-role norms shape LLM inference.
Paper + dataset: github.com/dhfbk/MRNI-BiasBe…
Yet another opportunity to party: congratulations to Davide Vandelli, who is currently a junior researcher in our group and just got his Master's degree in Data Science at @DS_UNITN on video game features to predict cyberbullying in young adolescents. Congrats!! 🥳🥳🥳
Congratulations to Dr. @paccosi_teresa who successfully defended (cum laude) her PhD thesis on frame-based approaches to trace sensory events in texts!! 🥳🥳🥳
Happy to share that “Fine-grained Fallacy Detection with Human Label Variation” with Agnese Daffara and @sara_hlt was accepted to #NAACL2025 main conference 🎉
📝 arxiv.org/abs/2502.13853
#NLProc #NLP #NAACL
1/🧵
DH Group at FBK retweeted
Arriva #Sanremo2025 e arriva anche il thread sull’analisi linguistica dei testi delle canzoni! #NLProc 🧵
Poster session is over. It was nice to chat about modality and factuality with many #coling2025 attendees, thanks for all the feedback! #NLProc
Good news! Our paper "Fine-grained Fallacy Detection with Human Label Variation" by @alanramponi, Agnese Daffara and @sara_hlt has been accepted at #NAACL2025 🥳 See you in New Mexico!! #NLProc
DH Group at FBK retweeted
Happy to see @itspalomap paper "Decoding Hate: Exploring Language Models' Reactions to Hate Speech" accepted at main conference at #NAACL2025 🎉📚✨ See you in Albuquerque🇺🇸! @HYBRIDS_MSCA
If you are attending #COLING2025 this week, don't miss Marco Rovera's poster presentation on Friday about modality and factuality annotation and detection in Italian! Paper here: aclanthology.org/2025.coling… #AICODEproject #NLProc
DH Group at FBK retweeted
MaiNLP works on dialects presented this week at COLING #NLProc 🥳
- A new cross-dialect IR benchmark (presented Wed, first session)
- testing PIXEL models on non-standard languages (Thu first session)
- Multi-dialectal Bavarian slot and intent detection and our participation in the Norwegian dialect shared task (VarDial)
see more in the 🧵
Replying to @CisLmu
📝Evaluating Pixel Language Models on Non-Standardized Languages
🔎We trained a German PIXEL model -- which performs strongly on dialect data
👥Alberto Muñoz-Ortiz, Verena Blaschke, @barbara_plank
🔗arxiv.org/abs/2412.09084
📁Main - Short
DH Group at FBK retweeted
We have an incredible lineup of speakers and panelists! 🎤 @malihealikhani @IAugenstein @dirk_hovy @ZhijingJin @vinodkpg @ChrisEmezue @nedjmaou @DZytko
DH Group at FBK retweeted
📢 Have you been wondering what workshops are brewing in the *ACL venues in 2025? The list that we've been waiting for in here. Feel free to tag or repost with the organisers. Below are ACL 2025 workshops: #ACL2025NLP #NLProc #workshop 🧵
DH Group at FBK retweeted
📢#jobalert Are you looking for a new challenge in 2025? Come work with me! I have a 3-year position for a postdoc on the intersection of #semanticweb, #maritimehistory and #nlproc. The successful candidate would join us in this historical building in downtown Amsterdam...1/6
Wishing everybody happy holidays before a well-deserved break! Greetings from our traditional Christmas lunch 🎄🧑🎄#NLProc #Digitalhumanities




















