PhD Student in Natural Language Processing at the University of Pisa #NLPROC
Joined January 2018
- Tweets 122
- Following 291
- Followers 52
- Likes 298
2 Photos and videos
Michele Papucci retweeted
May 14
🧶 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
1
3
10
457
Michele Papucci retweeted
May 11
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
1
2
115
Michele Papucci retweeted
May 11
🚀 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.
1
3
13
373
4 Aug 2025
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…
5
42
Michele Papucci retweeted
22 Jul 2025
🎉 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
1
4
6
503
Michele Papucci retweeted
20 Jun 2025
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
4
12
552
3 Jun 2025
🧵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
2
3
7
334
3 Jun 2025
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.
1
1
72
3 Jun 2025
8/ TL:DR;
🚨 State-of-the-art Detectors today are too shallow
📉 A bit of style alignment makes them crumble
🧠 We need stronger benchmarks
🛠 We develop a way to create hard, in-domain texts for making and evaluating the next generation of more robust and reliable MGT Detectors
1
61
Michele Papucci retweeted
29 May 2025
@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
29 May 2025
Our interpretability team recently released research that traced the thoughts of a large language model.
Now we’re open-sourcing the method. Researchers can generate “attribution graphs” like those in our study, and explore them interactively.
8
45
215
63,333
Michele Papucci retweeted
16 May 2025
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.)
1
1
5
207
Michele Papucci retweeted
16 May 2025
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 👇
1
3
9
887
Michele Papucci retweeted
5 Dec 2024
And now is the turn of Chiara Fazzone, presenting “SimilEx: the First Italian Dataset for Sentence Similarity with Natural Language Explanations” (with @AlzettaChiara, Dell’Orletta F. and Venturi G.)! 🔥 @CLiC_it_conf @AILC_NLP #NLProc
1
2
12
490
Michele Papucci retweeted
2 Dec 2024
[1/4] 🎉 Excited to announce that our paper "🥞TEXT-CAKE: Challenging Language Models on Local Text Coherence" (@lucadini_, F. Dell'Orletta, D. Brunato, @tommaso_caselli) has been accepted at COLING 2025 (@coling2025)! 📝✨ #NLProc #COLING2025
1
3
15
783