Overview
The Multimodal Algorithmic Reasoning Workshop (MAR) at NeurIPS 2025 invites researchers to submit their work on the intersection of multimodal learning and algorithmic reasoning. This workshop, now in its fifth iteration, will take place in San Diego on December 6 or 7, 2025. It aims to explore the advancements and challenges in multimodal AI, particularly in relation to human-like reasoning capabilities.
Background & Relevance
Multimodal learning integrates information from various sources, such as text, images, and audio, to enhance reasoning processes. This area is critical as AI systems increasingly demonstrate capabilities that may rival human intelligence. However, the question remains whether these models genuinely understand the tasks they perform or if they merely mimic human reasoning without comprehension. Addressing these issues is essential for the development of more robust AI systems that can solve complex problems effectively.
Key Details
- Event: Multimodal Algorithmic Reasoning Workshop (MAR) at NeurIPS 2025
- Dates: December 6 or 7, 2025
- Location: San Diego, California
- Submission Deadline: August 31, 2025 (Anywhere on Earth)
- Notification of Acceptance: September 22, 2025
- Camera Ready Submission: November 3, 2025 (tentative)
- Workshop Website: MAR Workshop
Eligibility & Participation
This workshop targets researchers and practitioners working in areas related to multimodal learning, mathematical reasoning, and large language models. Participants are encouraged to submit original and unpublished research that contributes to the understanding of multimodal algorithmic reasoning.
Submission or Application Guidelines
- Submissions must be original and previously unpublished.
- All papers should be submitted via the workshop’s CMT website: CMT Submission.
- Papers should be in PDF format and adhere to the provided submission style, which can be found here.
- The maximum length is four pages, excluding references.
- An optional appendix may be submitted with additional materials, with a maximum size of 50MB.
- All submissions must maintain author anonymity and follow NeurIPS double-blind review guidelines.
- Accepted papers will be presented at the workshop, and at least one author must attend in person.
- Accepted works will be publicly accessible on the workshop website after the camera-ready deadline, but will not be included in archival proceedings.
More Information
The workshop will feature keynote talks from prominent researchers in the field, including Yu Cheng, Noah Goodman, Kristen Grauman, Subbarao Kambhampati, and Max Tegmark. Topics of interest include multimodal algorithmic reasoning, comparisons between AI and human problem-solving, and the exploration of new architectures for visual reasoning. This workshop represents a significant opportunity for researchers to engage with current challenges and advancements in AI, particularly in understanding the limits and potentials of reasoning in large models.
Conclusion
Researchers are encouraged to explore this opportunity to contribute to the evolving discourse on multimodal algorithmic reasoning. By submitting your work, you can help shape the future of AI and its applications in complex problem-solving scenarios. For more details, visit the workshop website and prepare your submissions by the deadline.
Category: CFP & Deadlines
Tags: multimodal ai, algorithmic reasoning, neural processing, cognitive science, machine learning, deep learning, vision-language models, human problem solving, AI applications, NeurIPS, mathematical reasoning, distributed problem solving
