processing natural language to tackle social and economic challenges @Unibocconi @MilaNLProc
Joined January 2022
- Tweets 24
- Following 49
- Followers 35
- Likes 179
Photos and videos
Lorenzo Lupo retweeted
28 Oct 2025
Can AI simulate human behavior? 🧠
The promise is revolutionary for science & policy. But there’s a huge "IF": Do these simulations actually reflect reality?
To find out, we introduce SimBench: The first large-scale benchmark for group-level social simulation. (1/9)
2
23
53
5,595
23 Feb 2024
New paper accepted at @LrecColing 🥳
Check it out: lorelupo.github.io/docs/dadi…
23 Feb 2024
✨"DADIT: A Dataset for Demographic Classification of Italian Twitter Users" contains 20k Italian Twitter users labeled with their demographics and their content. The demographic classifiers trained on DADIT outperform SOTA like M3 thanks to the use of tweets.
1
2
7
583
12 Mar 2024
🎊 New paper accepted at @LrecColing!
DADIT contains 20k Italian Twitter users and their content, labelled with their demographics.
Demographic classifiers trained on DADIT significantly outperform popular classifiers like M3 thanks to the use of tweets.
arxiv.org/abs/2403.05700
1
7
26
4,272
12 Mar 2024
It was a pleasure to work with great co-authors @paulromanbose @MahyarHabibi007 @dirk_hovy and Carlo Schwarz @MilaNLProc
1
3
87
Lorenzo Lupo retweeted
11 Mar 2024
Very interesting dataset:
11 Mar 2024
#MemoryMonday #NLProc
"DADIT: A Dataset for Demographic Classification of Italian Twitter Users and a Comparison of Prediction Methods" by Lupo et al. presents a 30M tweet corpus from 20k Italian users with demographics and compares prediction methods.
lorelupo.github.io/docs/dadi…
1
3
5
727
Lorenzo Lupo retweeted
27 Feb 2024
There are many concerns around LLMs being politically biased. But how, if at all, can we meaningfully evaluate values/opinions in LLMs?
In our new paper, we show that current *constrained* evals (e.g. surveys) likely tell us little about LLM values/opinions in the real world
🧵
ALT A model is prompted with a proposition from the Political Compass Test. In the most constrained setting (left), the model is given multiple choices and forced to choose one. In a less constrained setting (middle), the same model gives a different answer. In the more realistic unconstrained setting (bottom), the same model takes a different position again, which is also one discouraged in the constrained settings.
5
26
109
20,565
Lorenzo Lupo retweeted
23 Feb 2024
📜The @MilaNLProc lab has 5 papers accepted at the upcoming conferences (@LrecColing, @eaclmeeting, @iclr_conf).
Here's a tl;dr for each of them ⤵️ 🧵
#LRECCOLING2024 #EACL2024 #ICLR2024 #NLProc
1
5
29
5,521
Lorenzo Lupo retweeted
20 Feb 2024
📆Reminder📆: The #WOAH2024 submission deadline is March 10, anywhere on Earth!
See CFP: workshopononlineabuse.com/cf…
#NAACL2024 #NLProc
5
8
1,373
Lorenzo Lupo retweeted
20 Feb 2024
Indeed, @SafeConvAI 2024 kindly welcomes safety-related work, consider submitting! 🚀
sites.google.com/view/safety…
20 Feb 2024
If you woke up to a rejection from @LRECCOLING, have some coffee, a good cry, and consider submitting @SafeConvAI Workshop! 👀
3
9
952
Lorenzo Lupo retweeted
22 Feb 2024
For this week's @MilaNLProc reading group,
@lorelupo presented "Value Kaleidoscope: Engaging AI with Pluralistic Human Values, Rights, and Duties" by @ma_tay_ et al.
Paper: arxiv.org/abs/2309.00779
#NLProc #ReadingGroup
6
14
955
Lorenzo Lupo retweeted
1 Feb 2024
For this week's @MilaNLProc reading group, @CurriedAmanda presented different papers related to AI regulation, with a focus on the EU AI Act.
1️⃣ napier-repository.worktribe.…
2️⃣ aclanthology.org/2023.emnlp-…
3️⃣ europarl.europa.eu/news/en/h…
#NLProc #ReadingGroup
ALT By @CurriedAmanda
3
14
1,476
Lorenzo Lupo retweeted
17 Jan 2024
🚨 Friendly Reminder 🚨
⏰ Time is running out! The deadline for your research visit application at @MilaNLProc is January 31. Don't miss this great opportunity to join us!
📝 Apply here: forms.gle/UVJY5QZm48H1fFZD8
#ResearchOpportunity #NLProc
8
28
8,853
Lorenzo Lupo retweeted
4 Jan 2024
If you’re working on LLM safety, check out SafetyPrompts.com!
SafetyPrompts.com is a catalogue of open datasets for evaluating and improving LLM safety. I started building this over the holidays, and I know there are still datasets missing, so I need your help 🧵
8
52
213
36,160
Lorenzo Lupo retweeted
21 Apr 2023
🎯 Do speak English and Italian?
🎯 Interested in informing research on inclusive language technologies?
🎯 Do you have 10 minutes?
Help us out by taking part in this *survey* on gender-neutral (Machine) translation!
Link --> forms.gle/twaJBhYMpT5pvG1t6
@fbk_mt
1
9
16
1,791
27 Apr 2023
So thrilled to work with this amazing team!
21 Apr 2023
🎉 Excited to welcome @lorelupo to our lab! His expertise in NLP, neural machine translation and reinforcement learning will bring new perspectives to our research.
#NLP #MachineTranslation #ML
milanlproc.github.io/authors…
1
4
629