Instruction

Large language models (LLMs) have achieved remarkable breakthroughs in natural language understanding and generation, but their logical reasoning capabilities remain a significant bottleneck. Logical reasoning is crucial for tasks requiring precise deduction, induction, or abduction, such as medical diagnosis, legal reasoning, and scientific hypothesis verification. However, LLMs often fail to handle complex logical problems with multiple premises and constraints, and they frequently produce self-contradictory responses across different questions. For example, a state-of-the-art Macaw question answering LLM answers Yes to both questions Is a magpie a bird? and Does a bird have wings?, but answers No to Does a magpie have wings?. These limitations not only restrict the reliability of LLMs in high-stakes scenarios but also hinder their ability to perform multi-step reasoning required for complex problem-solving. Addressing these challenges is essential to bridge the gap between current LLMs and the demands of real-world applications.

To facilitate this research direction of empowering LLMs with logical reasoning, our workshop aims to bring together researchers and practitioners from logic, artificial intelligence, natural language processing, and related fields to systematically address the core challenges in enhancing LLMs' logical reasoning abilities, sharing cutting-edge research progress and fostering interdisciplinary discussions on future directions. Specifically, the workshop will mainly focus on two primary aspects: logical question answering and logical consistency. For logical question answering, we will explore solver-based methods (e.g., translating natural language problems into symbolic logic for external solvers), prompt-based approaches (e.g., explicit logical chain modeling via Chain-of-Thought or symbolic reasoning through carefully designed prompts), and pretraining/fine-tuning techniques (e.g., augmenting training data with logical deduction processes). For logical consistency, we will discuss strategies to enhance negation consistency, implication consistency, transitivity consistency, factuality consistency, and compositional consistency, using methodologies spanning from neuro-symbolic integration, memory-augmented models, and constraint-based training.

Reviewer Recruitment

We are actively seeking reviewers to join our program committee. If you are interested in contributing to the peer-review process, please contact us via ijcai26-logisymb@googlegroups.com.

Topics

Topics of interest include, but are not limited to:

• • Logical question answering of LLMs;

  • Chain-of-thought reasoning (by explicitly modeling the logical reasoning chain) of LLMs;

  • External tool-use (e.g., logic solver) for LLMs' reasoning;

  • Logical consistency (e.g., implication consistency, negation consistency) of LLMs;

  • Mathematical and Symbolic reasoning (e.g., proof writing) via logical rules using LLMs.

Important Dates

• Submission Deadline: May 31, 2026

• Notification of Acceptance: June 14, 2026 

All deadlines are specified in Anywhere on Earth (AoE).

Submission Guidelines

We welcome two types of papers:

• Regular papers: Full-length research papers from 4 to 10 pages (excluding references and supplementary material);

• Tiny/Short papers: research/position papers from 2 to 4 pages (excluding references and supplementary material).

Papers should be submitted in the IJCAI 2026 format (https://www.ijcai.org/authors_kit). 

The review process will be double-blinded; therefore, all identifying information, including author names, affiliations, and acknowledgments, must be excluded from submissions. We welcome both unpublished manuscripts and previously accepted or published papers, provided that their submission does not violate any dual-submission policies. The contributions are non-archival. 

There will be Oral Presentation Awards, Outstanding Paper Awards, and one Best Paper Award for accepted papers with outstanding quality.

Submission Site

The workshop uses OpenReview for paper submission and reviewing. The submission link is https://openreview.net/group?id=ijcai.org/IJCAI-ECAI/2026/Workshop/LogiSymb.