Daniel Scalena

Milan, Italy

Hi! I am Daniel, a double PhD student at the 🇮🇹 University of Milano - Bicocca and the 🇳🇱 University of Groningen working on interpretability, fairness and security of generative (and non-generative) Large Language Models. My supervisors are Elisabetta Fersini and Malvina Nissim.

My research focuses on the use of interpretability as a tool to make generative models safer, more reliable and less toxic in order to extend and improve their real-world applications.

news

Dec 7, 2023	Presenting my poster and paper at the BlackBoxNLP workshop @EMNLP 2023 in Singapore 🇸🇬
Oct 26, 2023	Graduated! Thesis here 🎓
Jul 10, 2023	RewardLM (alpha) is now public! 🚀

latest posts

Sep 10, 2023	In-depth Notes to understand more about Transformers
Jul 12, 2023	Let the Models Respond: Interpreting the Detoxification process of LMs

selected publications

Let the Models Respond: Interpreting Language Model Detoxification Through the Lens of Prompt Dependence

Daniel Scalena, Gabriele Sarti, Malvina Nissim, and 1 more author

2023

Abs SOURCE PDF

Due to language models’ propensity to generate toxic or hateful responses, several techniques were developed to align model generations with users’ preferences. Despite the effectiveness of such methods in improving the safety of model interactions, their impact on models’ internal processes is still poorly understood. In this work, we apply popular detoxification approaches to several language models and quantify their impact on the resulting models’ prompt dependence using feature attribution methods. We evaluate the effectiveness of counter-narrative fine-tuning and compare it with reinforcement learning-driven detoxification, observing differences in prompt reliance between the two methods despite their similar detoxification performances.
MIND at SemEval-2023 Task 11: From Uncertain Predictions to Subjective Disagreement

Giulia Rizzi, Alessandro Astorino, Daniel Scalena, and 2 more authors

In Proceedings of the The 17th International Workshop on Semantic Evaluation (SemEval-2023), Jul 2023

Abs SOURCE PDF

This paper describes the participation of the research laboratory MIND, at the University of Milano-Bicocca, in the SemEval 2023 task related to Learning With Disagreements (Le-Wi-Di). The main goal is to identify the level of agreement/disagreement from a collection of textual datasets with different characteristics in terms of style, language and task.The proposed approach is grounded on the hypothesis that the disagreement between annotators could be grasped by the uncertainty that a model, based on several linguistic characteristics, could have on the prediction of a given gold label.