Zobrazeno 1 - 10
of 8 247
pro vyhledávání: '"Abate, P"'
Autor:
Abate, Nicolás, Torroba, Gonzalo
Information-theoretic methods have led to significant advances in nonperturbative quantum field theory in flat space. In this work, we show that these ideas can be generalized to field theories in a fixed de Sitter space. Focusing on 1+1-dimensional
Externí odkaz:
http://arxiv.org/abs/2409.18186
To evaluate the safety and usefulness of deployment protocols for untrusted AIs, AI Control uses a red-teaming exercise played between a protocol designer and an adversary. This paper introduces AI-Control Games, a formal decision-making model of the
Externí odkaz:
http://arxiv.org/abs/2409.07985
Autor:
Abate, Carlo, Bianchi, Filippo Maria
We propose a novel approach to compute the MAXCUT in attributed graphs, i.e., graphs with features associated with nodes and edges. Our approach is robust to the underlying graph topology and is fully differentiable, making it possible to find soluti
Externí odkaz:
http://arxiv.org/abs/2409.05100
Autor:
Voria, Gianmario, Sellitto, Giulia, Ferrara, Carmine, Abate, Francesco, De Lucia, Andrea, Ferrucci, Filomena, Catolino, Gemma, Palomba, Fabio
Machine learning's widespread adoption in decision-making processes raises concerns about fairness, particularly regarding the treatment of sensitive features and potential discrimination against minorities. The software engineering community has res
Externí odkaz:
http://arxiv.org/abs/2408.16683
Autor:
Benjamin, Patrick, Abate, Alessandro
Recent works have provided algorithms by which decentralised agents, which may be connected via a communication network, can learn equilibria in Mean-Field Games from a single, non-episodic run of the empirical system. However, these algorithms are g
Externí odkaz:
http://arxiv.org/abs/2408.11607
We present a data-driven approach for learning MDP policies that are robust across stochastic environments whose transition probabilities are defined by parameters with an unknown distribution. We produce probably approximately correct (PAC) guarante
Externí odkaz:
http://arxiv.org/abs/2408.03093
The goal of Bayesian inverse reinforcement learning (IRL) is recovering a posterior distribution over reward functions using a set of demonstrations from an expert optimizing for a reward unknown to the learner. The resulting posterior over rewards c
Externí odkaz:
http://arxiv.org/abs/2407.10971
In reinforcement learning, specifying reward functions that capture the intended task can be very challenging. Reward learning aims to address this issue by learning the reward function. However, a learned reward model may have a low error on the tra
Externí odkaz:
http://arxiv.org/abs/2406.15753
Autor:
Abate, Marco, Rakhimov, Karim
In this paper, we study the dynamics of geodesics of Fuchsian meromorphic connections with real periods, giving a precise characterization of the possible $\omega$-limit sets of simple geodesics in this case. The main tools are the study of the singu
Externí odkaz:
http://arxiv.org/abs/2406.13353
Leveraging human preferences for steering the behavior of Large Language Models (LLMs) has demonstrated notable success in recent years. Nonetheless, data selection and labeling are still a bottleneck for these systems, particularly at large scale. H
Externí odkaz:
http://arxiv.org/abs/2406.10023