Giuseppe Magazzù's picture

Giuseppe Magazzù

saiteki-kai

·

https://www.unimib.it/giuseppe-magazzu

AI & ML interests

My research focuses on the developement of safety mitigation strategies and benchmarks for large language models.

Recent Activity

liked a model about 1 hour ago

allenai/wildguard

liked a dataset about 1 month ago

liyang-ict/FineHarm

liked a dataset about 1 month ago

walledai/XSTest

View all activity

Organizations

upvoted an article 2 months ago

Article

MiniGuard-v0.1: Prem's Guardrail Model Redefining the Pareto Frontier

Dec 12, 2025

•

22

upvoted a changelog 2 months ago

Hugging Face Changelog

Community Evals and Benchmark Repositories

Feb 5

• 77

upvoted a paper 3 months ago

The Art of Saying No: Contextual Noncompliance in Language Models

Paper • 2407.12043 • Published Jul 2, 2024 • 5

upvoted a paper 5 months ago

Emergent Misalignment via In-Context Learning: Narrow in-context examples can produce broadly misaligned LLMs

Paper • 2510.11288 • Published Oct 13, 2025 • 48

upvoted 4 collections about 1 year ago

Llama 4

Llama 4 release • 13 items • Updated Apr 29, 2025 • 728

MetricX-24

A collection of MetricX-24 models (https://aclanthology.org/2024.wmt-1.35/) • 6 items • Updated Mar 12 • 12

Gemma 3 Release

28 items • Updated Mar 12 • 635

Minerva LLMs

The first family of LLMs pretrained from scratch on Italian. • 6 items • Updated Dec 7, 2024 • 40

upvoted 2 collections over 1 year ago

Quality Estimation

SOTA Machine Translation Quality Estimation models • 5 items • Updated Jan 10, 2024 • 5

Tower

Model weights and SFT data for Tower. • 11 items • Updated Nov 15, 2024 • 32