PhD Student in Natural Language Processing at the University of Pisa #NLPROC
Italia
Joined January 2018
- Tweets124
- Following291
- Followers53
- Likes300
Michele Papucci retweeted
🔥 @lucadini_ and @workerplacemint won the Best Student Paper Award at @CLiC_it_conf for their paper “From Weights to Representations: How Pre-Training Data Order Shapes Language Models”! 🎉🎉 @AILC_NLP #NLPRoc
Michele Papucci retweeted
🧶 1/3
We’re excited to share our second paper presented at #LREC2026 in Palma: “Controllable Sentence Simplification in Italian: Fine-Tuning Large Language Models on Automatically Generated Resources” by @mpapucci_, Giulia Venturi and Felice Dell'Orletta
Michele Papucci retweeted
In this work we present UGLD, a decoding method that nudges LLMs toward or against a predefined vocabulary at inference time. No fine-tuning needed and without loss of fluency.
📄 Paper: minifyurl.in/xGNf
🔧 GitHub: github.com/michelepapucci/ug…
🐍 Install UGLD: pip install ugld
Michele Papucci retweeted
🚀 Excited to share our latest work at READIxTSAR workshop at #LREC2026
"Lexical Conditioning of Model's Distribution through Uncertainty-gated Soft-Mixing of Probabilities" by @mpapucci_ Giulia Venturi and Felice Dell'Orletta.
Last week I presented my latest work at the Findings Poster Session of ACL 2025 in Vienna!
If you missed it, check it out 💥
michelepapucci.github.io/blo…
Michele Papucci retweeted
🎉 Great news! We got 9 papers accepted at CLiC-it 2025! Looking forward to presenting them this year in Cagliari! 🇮🇹
#CLiCit2025 #NLProc @CLiC_it_conf @AILC_NLP
Michele Papucci retweeted
We are at the Lectures on Computational Linguistics in Milan organized by @AILC_NLP ! 🔥 @mpapucci_ and @workerplacemint are presenting their PhD work during the poster session! 🥳 #NLProc
🧵1/ Machine-Generated Text (MGT) detection is failing.
Our paper, accepted at Findings of ACL 2025, shows that LLMs can easily fool generated-text detectors.
arxiv.org/abs/2505.24523
@pdrndr, Cristiano Ciaccio, @AlessioMiaschi, @gpuccetti92, Felice dell'Orletta, Andrea Esuli
7/ What about Humans?
Human performance was unaffected: they performed poorly in detecting machine-generated text (around 50% accuracy in a binary task) both before and after our alignment.
Michele Papucci retweeted
@mntssys and I are excited to announce circuit-tracer, a library that makes circuit-finding simple!
Just type in a sentence, and get out a circuit showing (some of) the features your model uses to predict the next token. Try it on @neuronpedia: shorturl.at/SUX2A
Michele Papucci retweeted
4) Findings: Stress-testing Machine Generated Text Detection: Shifting Language Models Writing Style to Fool Detectors (@pdrndr, @mpapucci_, Ciaccio C., @AlessioMiaschi, @gpuccetti92, Dell'Orletta F. and Esuli A.)
Michele Papucci retweeted
Exciting news!🔥 We got 4 papers accepted at #ACL2025NLP: 2 at the main conference and 2 in the Findings! See you in Vienna! 🎉 @aclmeeting #NLPRoc
Details in the thread 👇
