11 citations · 28 across the 14 of their papers we have counts for
3 papers · 1 filter
Graph-Regularized Sparse Autoencoders for LLM Safety Steering
Jehyeok Yeon, Federico Cinus, Yifan Wu +1
Sparse autoencoders (SAEs) are increasingly used to extract activation directions for inference-time steering, but their standard sparsity objective treats latent features as indep…
Online Minimization of Polarization and Disagreement via Low-Rank Matrix Bandits
Federico Cinus, Yuko Kuroki, Atsushi Miyauchi +1
We study the problem of minimizing polarization and disagreement in the Friedkin-Johnsen opinion dynamics model under incomplete information. Unlike prior work that assumes a stati…
Engagement-Driven Content Generation with Large Language Models
Erica Coppolillo, Federico Cinus, Marco Minici +2
Large Language Models (LLMs) demonstrate significant persuasive capabilities in one-on-one interactions, but their influence within social networks, where interconnected users and…