Autoencoders have been used for finding interpretable and disentangled features underlying neural network representations in both image and text domains. While the efficacy and pitfalls of such methods are well-studied in vision, there is a lack of corresponding results, both qualitative and quantitative, for the text domain. We aim to address this gap by training sparse autoencoders (SAEs) on a synthetic testbed of formal languages. Specifically, we train SAEs on the hidden representations of models trained on formal languages (Dyck-2, Expr, and English PCFG) under a wide variety of hyperparameter settings, finding interpretable latents often emerge in the features learned by our SAEs. However, similar to vision, we find performance turns out to be highly sensitive to inductive biases of the training pipeline. Moreover, we show latents correlating to certain features of the input do not always induce a causal impact on model's computation. We thus argue that causality has to become a central target in SAE training: learning of causal features should be incentivized from the ground-up. Motivated by this, we propose and perform preliminary investigations for an approach that promotes learning of causally relevant features in our formal language setting.

This paper presents a hybrid methodology that enhances the training process of deep learning (DL) models by embedding domain expert knowledge using ontologies and answer set programming (ASP). By integrating these symbolic AI methods, we encode domain-specific constraints, rules, and logical reasoning directly into the model's learning process, thereby improving both performance and trustworthiness. The proposed approach is flexible and applicable to both regression and classification tasks, demonstrating generalizability across various fields such as healthcare, autonomous systems, engineering, and battery manufacturing applications. Unlike other state-of-the-art methods, the strength of our approach lies in its scalability across different domains. The design allows for the automation of the loss function by simply updating the ASP rules, making the system highly scalable and user-friendly. This facilitates seamless adaptation to new domains without significant redesign, offering a practical solution for integrating expert knowledge into DL models in industrial settings such as battery manufacturing.

The problem of checking satisfiability of linear real arithmetic (LRA) and non-linear real arithmetic (NRA) formulas has broad applications, in particular, they are at the heart of logic-related applications such as logic for artificial intelligence, program analysis, etc. While there has been much work on checking satisfiability of unquantified LRA and NRA formulas, the problem of checking satisfiability of quantified LRA and NRA formulas remains a significant challenge. The main bottleneck in the existing methods is a computationally expensive quantifier elimination step. In this work, we propose a novel method for efficient quantifier elimination in quantified LRA and NRA formulas. We propose a template-based Skolemization approach, where we automatically synthesize linear/polynomial Skolem functions in order to eliminate quantifiers in the formula. The key technical ingredients in our approach are Positivstellens\"atze theorems from algebraic geometry, which allow for an efficient manipulation of polynomial inequalities. Our method offers a range of appealing theoretical properties combined with a strong practical performance. On the theory side, our method is sound, semi-complete, and runs in subexponential time and polynomial space, as opposed to existing sound and complete quantifier elimination methods that run in doubly-exponential time and at least exponential space. On the practical side, our experiments show superior performance compared to state-of-the-art SMT solvers in terms of the number of solved instances and runtime, both on LRA and on NRA benchmarks.

PyPM is a Python-based domain specific language (DSL) for building rewrite-based optimization passes on machine learning computation graphs. Users define individual optimizations by writing (a) patterns that match subgraphs of a computation graph and (b) corresponding rules which replace a matched subgraph with an optimized kernel. PyPM is distinguished from the many other DSLs for defining rewriting passes by its complex and novel pattern language which borrows concepts from logic programming. PyPM patterns can be recursive, nondeterminstic, and can require checking domain-specific constraints such as the shapes of tensors. The PyPM implementation is thus similarly complicated, consisting of thousands of lines of C++ code. In this paper, we present our work on building PyPM, as well as formalizing and distilling and this complexity to an understandable mathematical core. We have developed a formal core calculus expressing the main operations of the PyPM pattern language. We define both a declarative semantics - describing which patterns match which terms - and an algorithmic semantics - an idealized version of the PyPM pattern interpreter - and prove their equivalence. The development is fully mechanized in the Coq proof assistant.

Sequential problems are ubiquitous in AI, such as in reinforcement learning or natural language processing. State-of-the-art deep sequential models, like transformers, excel in these settings but fail to guarantee the satisfaction of constraints necessary for trustworthy deployment. In contrast, neurosymbolic AI (NeSy) provides a sound formalism to enforce constraints in deep probabilistic models but scales exponentially on sequential problems. To overcome these limitations, we introduce relational neurosymbolic Markov models (NeSy-MMs), a new class of end-to-end differentiable sequential models that integrate and provably satisfy relational logical constraints. We propose a strategy for inference and learning that scales on sequential settings, and that combines approximate Bayesian inference, automated reasoning, and gradient estimation. Our experiments show that NeSy-MMs can solve problems beyond the current state-of-the-art in neurosymbolic AI and still provide strong guarantees with respect to desired properties. Moreover, we show that our models are more interpretable and that constraints can be adapted at test time to out-of-distribution scenarios.

In this paper, we introduce a syntactic framework for analyzing and handling inconsistencies in propositional bases. Our approach focuses on examining the relationships between variable occurrences within conflicts. We propose two dual concepts: Minimal Inconsistency Relation (MIR) and Maximal Consistency Relation (MCR). Each MIR is a minimal equivalence relation on variable occurrences that results in inconsistency, while each MCR is a maximal equivalence relation designed to prevent inconsistency. Notably, MIRs capture conflicts overlooked by minimal inconsistent subsets. Using MCRs, we develop a series of non-explosive inference relations. The main strategy involves restoring consistency by modifying the propositional base according to each MCR, followed by employing the classical inference relation to derive conclusions. Additionally, we propose an unusual semantics that assigns truth values to variable occurrences instead of the variables themselves. The associated inference relations are established through Boolean interpretations compatible with the occurrence-based models.

Recently, it is often said that the data used for the pre-training of large language models (LLMs) have been exhausted. This paper proposes a solution to the problem: Automated generation of massive reasonable empirical theorems by forward reasoning based on strong relevant logics. In fact, this can be regarded as a part of our approach to the problems of ATF (Automated Theorem Finding) and AKA (Automated Knowledge Appreciation).

This paper shows that the semantics of programs with aggregates implemented by the solvers clingo and dlv can be characterized as extended First-Order formulas with intensional functions in the logic of Here-and-There. Furthermore, this characterization can be used to study the strong equivalence of programs with aggregates under either semantics. We also present a transformation that reduces the task of checking strong equivalence to reasoning in classical First-Order logic, which serves as a foundation for automating this procedure.

Answer Set Programming (ASP), a well-known declarative logic programming paradigm, has recently found practical application in Process Mining. In particular, ASP has been used to model tasks involving declarative specifications of business processes. In this area, Declare stands out as the most widely adopted declarative process modeling language, offering a means to model processes through sets of constraints valid traces must satisfy, that can be expressed in Linear Temporal Logic over Finite Traces (LTLf). Existing ASP-based solutions encode Declare constraints by modeling the corresponding LTLf formula or its equivalent automaton which can be obtained using established techniques. In this paper, we introduce a novel encoding for Declare constraints that directly models their semantics as ASP rules, eliminating the need for intermediate representations. We assess the effectiveness of this novel approach on two Process Mining tasks by comparing it with alternative ASP encodings and a Python library for Declare. Under consideration in Theory and Practice of Logic Programming (TPLP).

In many situations humans have to reason with inconsistent knowledge. These inconsistencies may occur due to not fully reliable sources of information. In order to reason with inconsistent knowledge, it is not possible to view a set of premisses as absolute truths as is done in predicate logic. Viewing the set of premisses as a set of assumptions, however, it is possible to deduce useful conclusions from an inconsistent set of premisses. In this paper a logic for reasoning with inconsistent knowledge is described. This logic is a generalization of the work of N. Rescher [15]. In the logic a reliability relation is used to choose between incompatible assumptions. These choices are only made when a contradiction is derived. As long as no contradiction is derived, the knowledge is assumed to be consistent. This makes it possible to define an argumentation-based deduction process for the logic. For the logic a semantics based on the ideas of Y. Shoham [22, 23], is defined. It turns out that the semantics for the logic is a preferential semantics according to the definition S. Kraus, D. Lehmann and M. Magidor [12]. Therefore the logic is a logic of system P and possesses all the properties of an ideal non-monotonic logic.