Safe reinforcement learning (RL) is crucial for real-world applications, and multi-agent interactions introduce additional safety challenges. While Probabilistic Logic Shields (PLS) has been a powerful proposal to enforce safety in single-agent RL, their generalizability to multi-agent settings remains unexplored. In this paper, we address this gap by conducting extensive analyses of PLS within decentralized, multi-agent environments, and in doing so, propose Shielded Multi-Agent Reinforcement Learning (SMARL) as a general framework for steering MARL towards norm-compliant outcomes. Our key contributions are: (1) a novel Probabilistic Logic Temporal Difference (PLTD) update for shielded, independent Q-learning, which incorporates probabilistic constraints directly into the value update process; (2) a probabilistic logic policy gradient method for shielded PPO with formal safety guarantees for MARL; and (3) comprehensive evaluation across symmetric and asymmetrically shielded n-player game-theoretic benchmarks, demonstrating fewer constraint violations and significantly better cooperation under normative constraints. These results position SMARL as an effective mechanism for equilibrium selection, paving the way toward safer, socially aligned multi-agent systems.

Global cooperation on climate change mitigation is essential to limit temperature increases while supporting long-term, equitable economic growth and sustainable development. Achieving such cooperation among diverse regions, each with different incentives, in a dynamic environment shaped by complex geopolitical and economic factors, without a central authority, is a profoundly challenging game-theoretic problem. This article introduces RICE-N, a multi-region integrated assessment model that simulates the global climate, economy, and climate negotiations and agreements. RICE-N uses multi-agent reinforcement learning (MARL) to encourage agents to develop strategic behaviors based on the environmental dynamics and the actions of the others. We present two negotiation protocols: (1) Bilateral Negotiation, an exemplary protocol and (2) Basic Club, inspired from Climate Clubs and the carbon border adjustment mechanism (Nordhaus, 2015; Comissions, 2022). We compare their impact against a no-negotiation baseline with various mitigation strategies, showing that both protocols significantly reduce temperature growth at the cost of a minor drop in production while ensuring a more equitable distribution of the emission reduction costs.

Communication is a widely used mechanism to promote cooperation in multi-agent systems. In the field of emergent communication, agents are typically trained in specific environments: cooperative, competitive or mixed-motive. Motivated by the idea that real-world settings are characterized by incomplete information and that humans face daily interactions under a wide spectrum of incentives, we aim to explore the role of emergent communication when simultaneously exploited across all these contexts. In this work, we pursue this line of research by focusing on social dilemmas. To do this, we developed an extended version of the Public Goods Game, which allows us to train independent reinforcement learning agents simultaneously in different scenarios where incentives are (mis)aligned to various extents. Additionally, agents experience uncertainty in terms of the alignment of their incentives with those of others. We equip agents with the ability to learn a communication policy and study the impact of emergent communication in the face of uncertainty among agents. Our findings show that in settings where all agents have the same level of uncertainty, communication can enhance the cooperation of the whole group. However, in cases of asymmetric uncertainty, the agents that do not face uncertainty learn to use communication to deceive and exploit their uncertain peers.

Recent developments in face restoration have achieved remarkable results in producing high-quality and lifelike outputs. The stunning results however often fail to be faith-ful with respect to the identity of the person as the models lack necessary context. In this paper, we explore the poten-tial of personalized face restoration with diffusion models. In our approach a restoration model is personalized using a few images of the identity, leading to tailored restoration with respect to the identity while retaining fine-grained de-tails. By using independent trainable blocks for personal-ization, the rich prior of a base restoration model can be ex-ploited to its fullest. To avoid the model relying on parts of identity left in the conditioning low-quality images, a gener-ative regularizer is employed. With a learnable parameter, the model learns to balance between the details generated based on the input image and the degree of personalization. Moreover, we improve the training pipeline of face restoration models to enable an alignment-free approach. We showcase the robust capabilities of our approach in sev-eral real-world scenarios with multiple identities, demon-strating our method's ability to generate fine-grained de-tails with faithful restoration. In the user study we evalu-ate the perceptual quality and faithfulness of the generated details, with our method being voted best 61% of the time compared to the second best with 25% of the votes.

Addressing the question of how to achieve optimal decision-making under risk and uncertainty is crucial for enhancing the capabilities of artificial agents that collaborate with or support humans. In this work, we address this question in the context of Public Goods Games. We study learning in a novel multi-objective version of the Public Goods Game where agents have different risk preferences, by means of multi-objective reinforcement learning. We introduce a parametric non-linear utility function to model risk preferences at the level of individual agents, over the collective and individual reward components of the game. We study the interplay between such preference modelling and environmental uncertainty on the incentive alignment level in the game. We demonstrate how different combinations of individual preferences and environmental uncertainty sustain the emergence of cooperative patterns in non-cooperative environments (i.e., where competitive strategies are dominant), while others sustain competitive patterns in cooperative environments (i.e., where cooperative strategies are dominant).

Abstract Color constancy is still one of the biggest challenges in camera color processing. Convolutional neural networks have been able to improve the situation but there are still problems in many conditions, especially in scenes where a single color is dominating. In this work, we approach the problem from a slightly different setting. What if we could have some other information than the raw RGB image data. What kind of information would help to bring significant improvements while still be feasible in a mobile device. These questions sparked an idea for a novel approach for computational color constancy. Instead of raw RGB images used by the existing algorithms to estimate the scene white points, our approach is based on the scene’s average color spectra-single pixel spectral measurement. We show that as few as 10–14 spectral channels are sufficient. Notably, the sensor output has five orders of magnitude less data than in raw RGB images of a 10MPix camera. The spectral sensor captures the “spectral fingerprints” of different light sources and the illuminant white point can be accurately estimated by a standard regressor. The regressor can be trained with generated measurements using the existing RGB color constancy datasets. For this purpose, we propose a spectral data generation pipeline that can be used if the dataset camera model is known and thus its spectral characterization can be obtained. To verify the results with real data, we collected a real spectral dataset with a commercial spectrometer. On all datasets the proposed Single Pixel Spectral Color Constancy obtains the highest accuracy in the both single and cross-dataset experiments. The method is particularly effective for the difficult scenes for which the average improvements are 40–70% compared to state-of-the-arts. The approach can be extended to multi-illuminant case for which the experimental results also provide promising results.

For image segmentation, the current standard is to perform pixel-level optimization and inference in Euclidean output embedding spaces through linear hyperplanes. In this work, we show that hyperbolic manifolds provide a valuable alternative for image segmentation and propose a tractable formulation of hierarchical pixel-level classification in hyperbolic space. Hyperbolic Image Segmentation opens up new possibilities and practical benefits for segmentation, such as uncertainty estimation and boundary information for free, zero-label generalization, and increased performance in low-dimensional output embeddings.

The AI community is increasingly putting its attention towards combining symbolic and neural approaches, as it is often argued that the strengths and weaknesses of these approaches are complementary. One recent trend in the literature is weakly supervised learning techniques that employ operators from fuzzy logics. In particular, these use prior background knowledge described in such logics to help the training of a neural network from unlabeled and noisy data. By interpreting logical symbols using neural networks, this background knowledge can be added to regular loss functions, hence making reasoning a part of learning. We study, both formally and empirically, how a large collection of logical operators from the fuzzy logic literature behave in a differentiable learning setting. We find that many of these operators, including some of the most well-known, are highly unsuitable in this setting. A further finding concerns the treatment of implication in these fuzzy logics, and shows a strong imbalance between gradients driven by the antecedent and the consequent of the implication. Furthermore, we introduce a new family of fuzzy implications (called sigmoidal implications) to tackle this phenomenon. Finally, we empirically show that it is possible to use Differentiable Fuzzy Logics for semi-supervised learning, and compare how different operators behave in practice. We find that, to achieve the largest performance improvement over a supervised baseline, we have to resort to non-standard combinations of logical operators which perform well in learning, but no longer satisfy the usual logical laws.

Influence diagrams (IDs) are well-known formalisms, which extend Bayesian networks to model decision situations under uncertainty. Although they are convenient as a decision theoretic tool, their knowledge representation ability is limited in capturing other crucial notions such as logical consistency. In this article, we complement IDs with the light-weight description logic (DL) EL to overcome such limitations. We consider a setup where DL axioms hold in some contexts, yet the actual context is uncertain. The framework benefits from the convenience of using DL as a domain knowledge representation language and the modelling strength of IDs to deal with decisions over contexts in the presence of contextual uncertainty. We define related reasoning problems and study their computational complexity.

Computation Tree Logic (CTL) is one of the central formalisms in formal verification. As a specification language, it is used to express a property that the system at hand is expected to satisfy. From both the verification and the system design points of view, some information content of such property might become irrelevant for the system due to various reasons, e.g., it might become obsolete by time, or perhaps infeasible due to practical difficulties. Then, the problem arises on how to subtract such piece of information without altering the relevant system behaviour or violating the existing specifications over a given signature. Moreover, in such a scenario, two crucial notions are informative: the strongest necessary condition (SNC) and the weakest sufficient condition (WSC) of a given property. To address such a scenario in a principled way, we introduce a forgetting-based approach in CTL and show that it can be used to compute SNC and WSC of a property under a given model and over a given signature. We study its theoretical properties and also show that our notion of forgetting satisfies existing essential postulates of knowledge forgetting. Furthermore, we analyse the computational complexity of some basic reasoning tasks for the fragment CTLAF in particular.

Combining symbolic and neural approaches has gained considerable attention in the AI community, as it is argued that their strengths and weaknesses are complementary. One trend in the literature are weakly supervised learning techniques that employ operators from fuzzy logics. They use prior background knowledge described in such logics to help training neural networks from unlabeled and noisy data. By interpreting logical symbols using neural networks (or grounding them), this background knowledge can be added to regular loss functions, hence making reasoning a part of learning. We investigate how implications from the fuzzy logic literature behave in a differentiable setting. In this setting, we analyze the differences between the formal properties of these fuzzy implications. It turns out that various fuzzy implications, including some of the most well-known, are highly unsuitable for use in a differentiable learning setting. A further finding shows a strong imbalance between gradients driven by the antecedent and the consequent of the implication. Furthermore, we introduce a new family of fuzzy implications (called sigmoidal implications) to tackle this phenomenon. Finally, we empirically show that it is possible to use Differentiable Fuzzy Logics for semi-supervised learning, and show that sigmoidal implications outperform other choices of fuzzy implications.

In the design of complex systems, model-checking and satisfiability arise as two prominent decision problems. While model-checking requires the designed system to be provided in advance, satisfiability allows to check if such a system even exists. With very few exceptions, the second problem turns out to be harder than the first one from a complexity-theoretic standpoint. In this paper, we investigate the connection between the two problems for a non-trivial fragment of Strategy Logic (SL, for short). SL extends LTL with first-order quantifications over strategies, thus allowing to explicitly reason about the strategic abilities of agents in a multi-agent system. Satisfiability for the full logic is known to be highly undecidable, while model-checking is non-elementary.The SL fragment we consider is obtained by preventing strategic quantifications within the scope of temporal operators. The resulting logic is quite powerful, still allowing to express important game-theoretic properties of multi-agent systems, such as existence of Nash and immune equilibria, as well as to formalize the rational synthesis problem. We show that satisfiability for such a fragment is PSPACE-COMPLETE, while its model-checking complexity is 2EXPTIME-HARD. The result is obtained by means of an elegant encoding of the problem into the satisfiability of conjunctive-binding first-order logic, a recently discovered decidable fragment of first-order logic.

In a large and decentralised knowledge representation system such as the Web of Data, it is common for data sets to overlap. In the absence of a central naming authority, semantic heterogeneity is inevitable as such overlapping contents are described using different schemas. To overcome this problem, a number of solutions have automated the integration of these data sets by matching their schemas. In this work, we focus on a specific category of these solutions that relies on the concepts' extension for matching the schemas (i.e., instance-based methods). Rather than introducing a new approach for the task of schema matching, this work studies the impact of exploiting the semantics of owl:sameAs in such instance-based methods. For this empirical analysis, we investigate more than 900K concepts extracted from the Web, and make use of over 35B implicit identity assertions to study their impact. The experiments show that despite the growing doubts over their quality, exploiting owl:sameAs assertions extracted from the Web can improve instance-based schema matching techniques.

This paper introduces the Protégé plugin uDecide. With the help of uDecide it is possible to solve multi-attribute decision making problems encoded in a straight forward extension of standard Description Logics. The formalism allows to specify background knowledge in terms of an ontology, while each attribute is represented as a weighted class expression. On top of such an approach one can compute the best choice (or the best k-choices) taking background knowledge into account in the appropriate way. We show how to implement the approach on top of existing semantic web technologies and demonstrate its benefits with the help of an interesting use case that illustrates how to convert an existing web resource into an expert system with the help of uDecide.