计算机科学中的逻辑：2025–2026 年度综述

An annual survey of the cs.LO arXiv category, covering research published from September 2025 – August 2026. Built on a close-reading set of 108 papers and organised into 40 chapters, it maps the year by the field's research sub-directions.

Book framework (share of the year's output):

• Graph-Based Knowledge Representation and Neural-Symbolic Learning (87.2%) — This cluster focuses on integrating logical reasoning with graph-structured knowledge and neural-symbolic approaches for agents and applications.
• Interactive and Automated Theorem Proving in Proof Assistants (5.5%) — This cluster centers on developing and optimizing proof calculi, automated provers, and proof assistants like Lean and Isabelle for formal mathematics.
• Formal Verification of Temporal and Behavioral Properties (1.8%) — This cluster addresses verifying temporal, behavioral, and safety properties of complex programs and systems using formal semantics and model checking.
• Large Language Models for Logical Reasoning and Formalization (1.7%) — This cluster investigates the use of large language models to assist in natural language reasoning, formal proof generation, and structured logical tasks.
• SAT and Constraint Solving Techniques and Applications (1.4%) — This cluster focuses on advancing satisfiability and constraint solving methods, including encodings, parallel solving, and symbolic reasoning for practical problems.
• Dependent Type Theory and Categorical Semantics (1.4%) — This cluster explores theoretical foundations and semantic models of dependent type theories, including categorical and impredicative approaches.
• Fixed Points and Infinite-State System Verification (0.5%) — This cluster deals with fixed-point theory, modal mu-calculus, and verification techniques for infinite-state and real-time systems.
• Logic Education and Computational Infrastructure (0.3%) — This cluster addresses the development of computational tools, libraries, and educational resources for teaching and practicing logic in computer science.
• Statistical Model Checking and Topological Logics (0.2%) — This cluster focuses on statistical model checking techniques and topological logics for path-reachability and agent-based models.

Each sub-direction is covered across three chapters — how it framed its questions, the strength of its key findings, and its open debates and frontiers — so the reader gets a traceable picture of what this part of the field actually produced over the year.