
I'm a PhD in Data Science at @unitorvergata, now visiting @EdinburghUni. My research focuses on LLMs handling Formal Structures spanning from Logic to Code.
Frascati, Rome
Joined March 2023
- Tweets127
- Following184
- Followers90
- Likes491
LLMs may implicitly learn program structure.
🐴 BoJac leverages this knowledge for low-resource transpilation, inspired by the Böhm–Jacopini Theorem.
“Surfacing Program Structure with Auxiliary Signals for Low-Resource Code Translation”
Paper: aclanthology.org/2026.emnlp-…
#EMNLP2026
📢 Call for Papers: ORACLE – EMNLP 2026! 🇭🇺
Are #LLMs truly global oracles?
Translation is sorted, but what about reasoning? Can our models think and act across different cultures and languages? 🔮
📅 Dates
Deadline: 1/8/26
Workshop day: 28/10
Website: sites.google.com/view/worksh…
Fede_Ranaldi retweeted
📢 Call for Papers: ORACLE – EMNLP 2026! 🇭🇺
Are #LLMs truly global oracles?
Translation is sorted, but what about reasoning? Can our models think and act across different cultures and languages? 🔮
📅 Dates
Deadline: 1/8/26
Workshop day: 28/10/26
sites.google.com/view/worksh…
Fede_Ranaldi retweeted
Want to work on AI for code with one of the best NLP advisors in Europe?
PhD position open with Prof Shay Cohen at Edinburgh ILCC: AI4Code.
Deadline 10 May 2026.
edin.ac/4t4DRRL
Fede_Ranaldi retweeted
Can LLMs reliably predict program termination?
We evaluate frontier LLMs in the International Competition on Software Verification (SV-COMP) 2025, directly competing with state-of-the-art verification systems.
@AIatMeta @HebrewU @Bloomberg @imperialcollege @ucl
@jordiae @pascalkesseli @jvanegue @HyadataLab @adiyossLC @PeterOHearn12
Paper: arxiv.org/pdf/2601.18987
Website: orensultan.com/llms_halting_…
🧵👇
1/n
Fede_Ranaldi retweeted
🤔💭What even is reasoning? It's time to answer the hard questions!
We built the first unified taxonomy of 28 cognitive elements underlying reasoning
Spoiler—LLMs commonly employ sequential reasoning, rarely self-awareness, and often fail to use correct reasoning structures🧠
Fede_Ranaldi retweeted
🤔What happens when LLM agents choose between achieving their goals and avoiding harm to humans in realistic management scenarios? Are LLMs pragmatic or prefer to avoid human harm?
🚀 New paper out: ManagerBench: Evaluating the Safety-Pragmatism Trade-off in Autonomous LLMs🚀🧵
Fede_Ranaldi retweeted
Last but not least (for today), @lucadini_ presenting “The Role of Eye-Tracking Data in Encoder-Based Models: An In-depth Linguistic Analysis” (with @workerplacemint, Dominique Brunato and Felice Dell’Orletta)! 👁️
Link to the paper: clic2025.unica.it/wp-content…
#NLProc
I propose the concept of #protoknowledge (in-between #memorization and #generalization) to evaluate whether models are truly capable of solving downstream tasks. Thanks to the @l2m2_workshop part of @aclmeeting , I could receive valuable feedback and engage idea exchanges.
Fede_Ranaldi retweeted
Replying to @KeremZaman3
The poster will be presented at @l2m2_workshop – at Vienna, Austria
Fede_Ranaldi retweeted
💣💣This morning at #ACL2025, preventing these two brothers from fighting💣💣(they're R2, I am MultiOmnia, lol) #ACL2025NLP
#CompLing #Vienna
aclanthology.org/2025.acl-lo…
@l__ranaldi @FedeRanaldi
Fede_Ranaldi retweeted
Very excited to be here! @aclmeeting #ACL2025NLP #ACL2025
This morning presented our paper on Multimodal Multilingual Reasoning!
aclanthology.org/2025.acl-lo…
Many interesting interactions, feedback and new ideas for follow up!
Thank you guys @FedeRanaldi @Giuli12P2
Fede_Ranaldi retweeted
Willing to know how to remove Private Information from trained #LLMs ?
Private Memorization Editing is for you!
#ACL2025NLP #ACL2025
w/ @esruzzetti @commentino @znz8 – at Vienna, Austria
I will be at #ACL2025 with my group presenting 3 Conference Papers.
At the #L2M2 workshop, we will introduce the concept of #protoknowledge as a framework for jointly analyzing the #memorization and #generalization capabilities of LLMs.
Link Non-archival: lnkd.in/deDqJAxM
Privacy, Memorization, Multimodal reasoning, and the surge of protoknowledge (non-archival in L2M2 Workshop) ! This is our contribution to #ACL2025NLP to better understand #LLMs
We want to know your POV!
See you in Vienna!
We are hiring.
Fede_Ranaldi retweeted
Privacy, Memorization, Multimodal reasoning, and the surge of protoknowledge (non-archival in L2M2 Workshop) ! This is our contribution to #ACL2025NLP to better understand #LLMs
We want to know your POV!
See you in Vienna!
We are hiring.
Fede_Ranaldi retweeted
More info at:
humancentricart.github.io/
Fede_Ranaldi retweeted
Are you interested in the intersection of Mathematics and NLP?
Consider submitting your paper to #MathNLP 2025: The 3rd Workshop on Mathematical NLP. #EMNLP2025.
Submissions will open on June 25!
Take a look here for more details sites.google.com/view/mathnl…
Fede_Ranaldi retweeted
BlackboxNLP is back! 💥
Happy to be part of the organizing team for this year, and super excited for our new shared task using the excellent MIB Benchmark for circuit/causal variable localization in LMs, check it out! blackboxnlp.github.io/2025/t…
BlackboxNLP will be co-located with #EMNLP2025 in Suzhou this November! 📷This edition will feature a new shared task on circuits/causal variable localization in LMs, details: blackboxnlp.github.io/2025/t…
If you're into mech interp and care about evaluation, please submit!
Fede_Ranaldi retweeted
For the first time, researchers analyzed data contamination in LLM-aided RTL generation using established methods for contamination detection, showing that data contamination is a critical concern.
arxiv.org/pdf/2503.13572























