arXiv subjects

Thomas Pammer

Publications and source records attributed to Thomas Pammer.

3 recordsLinked to original sources

arXiv2026

From Natural Language to Exact Cover: A Neuro-Symbolic Approach to Zebra Puzzles

Chain-of-Thought (CoT) generation has substantially improved the performance of Large Language Models (LLMs) on complex reasoning tasks, including code generation, data analysis, and exam-style question answering. Despite these advances, purely neural LLMs continue to struggle with elementary logical reasoning problems and lack the determinism, soundness, and reliability characteristic of symbolic reasoning systems. Conversely, classical symbolic methods such as SAT solving and exact cover guarantee correctness and completeness, but require problems to be expressed in highly specialized formal encodings, limiting their applicability to natural language inputs. In this work, we present a synergistic neuro-symbolic framework that bridges this gap by combining neural semantic parsing with deterministic exact cover reasoning. Given a Zebra-style logic puzzle in natural language, an LLM translates entities and constraints into structured tool calls that construct an exact cover instance. A symbolic solver then computes the solution and provides feedback to iteratively refine invalid, under- or overconstrained encodings, exploiting the uniqueness property of Zebra puzzles. We conduct a comprehensive empirical evaluation across multiple parameter scales, post-training paradigms, and LLM families. The results demonstrate that our hybrid approach consistently outperforms strong plain neural baselines, including CoT prompting, as well as recent neuro-symbolic methods.

cs.AI↗

arXiv2026

Optimizing Sparse Boolean Rule Networks using Mixed Integer Linear Programming

We study the problem of learning accurate and cognitively simple binary classifiers that are directly interpretable as Boolean rules. We propose mixed integer linear programming (MILP) formulations for training sparse disjunctive normal form (DNF) networks, viewed as small networks with a conjunctive hidden layer and a disjunctive output layer. Contrary to previous work that uses MILP for inductive rule learning, we start with a fixed network structure, for which we then find optimal weights. A regularized objective explicitly trades off misclassification costs against sparsity, yielding globally optimized rule sets. Our main technical contribution is an enhanced formulation that substantially reduces the number of integer decision variables and eliminates redundant constraints, while preserving exact Boolean semantics for all feasible solutions. Furthermore, we develop a combined approach based on initialization and dropout to break symmetries in the weights between rules, mitigating redundancy, and encouraging more diverse and stable rule representations. Experiments on synthetic benchmarks and UCI datasets show that the advanced formulation improves computational efficiency and achieves competitive or superior accuracy compared to established interpretable baselines, including SAT-based DNF learning and well-known rule- and tree-based learners. Overall, our approach provides a principled optimization framework for learning compact Boolean rule sets with explicit control over accuracy and sparsity trade-offs.

cs.LG↗

arXiv2025

A matheuristic for solving the single row facility layout problem

The single row facility layout problem (SRFLP) is a well-studied NP-hard combinatorial optimization problem with applications in manufacturing and logistics systems. In the SRFLP, a set of facilities with lengths is given, as well as weights between each pair of facilities. The facilities must be arranged on a line, such that the sum of the weighted center-to-center distances is minimized. In this work, we introduce a novel matheuristic approach that integrates exact optimization into a metaheuristic framework based on simulated annealing to effectively solve large-scale SRFLP instances. Specifically, we propose the window approach matheuristic, which solves subsegments of the layout to optimality using mixed-integer programming while preserving the ordering of facilities outside the window. To the best of our knowledge, this constitutes the first matheuristic approach specifically designed for the SRFLP. We evaluate the performance of our method on the widely-used benchmark instance sets from literature. The computational results demonstrate that our matheuristic improves the best-known solution values for 17 of 70 instances, and matches the best-known solution values for the remaining 53 instances, outperforming current state-of-the-art metaheuristics.

math.OC↗