Neurosymbolic, or neural-symbolic, AI has emerged as a promising research direction, aiming to fuse the representational power of neural networks with the compositional and interpretable nature of symbolic reasoning. Such hybrid architectures promise bottom-up perceptual strength combined with top-down logical inference. Although the idea is not new, interest in neurosymbolic methods has surged recently. The renewed momentum stems partly from our improved understanding of neural architectures and the recognition that integrating symbolic reasoning may address some of the fundamental limitations of current deep-learning paradigms. Despite this momentum, neurosymbolic methods remain relatively unexplored within robotics. Yet robotics is uniquely positioned to benefit from such integration: robots process rich, continuous sensorimotor data streams but must also reason abstractly to plan, act safely, and achieve long-term goals. A neurosymbolic robotics framework could therefore unify low-level perception with high-level cognition, pushing the field closer to genuinely intelligent robotic systems.
This workshop seeks to highlight neurosymbolic approaches, explore their potential for cognitive robotics, and foster discussion on promising research directions. The organizers give special emphasis on developmental and bottom-up approaches, especially symbol emergence and rule learning from experience. We aim to spark debate on the feasibility and transformative potential of bringing recent neurosymbolic approaches into developmental robotics. Additionally, we would like to discuss current neuro-symbolic approaches in relation to foundational models, especially Visual Language Models and Visual Language Action Models, which we expect to offer new perspectives to the ICDL community.
We welcome submissions on related topics, including but not limited to the following:
We encourage researchers to submit 1-2 page extended abstracts or a preprint of relevant work. There is no page limitation for the preprint. Accepted abstracts will be posted on the workshop website, given the authors’ consent, and will not appear in the official IEEE proceedings. The review process is single-blind and will be conducted by the workshop organizers.
We also encourage researchers to submit their novel full-paper contributions. Accepted papers will be posted on the workshop website, given the authors’ consent, and will not appear in the official IEEE proceedings. They might also be presented as a short talk.
Please submit your contribution to emre.ugur@bogazici.edu.tr with the subject line: ICDL WS contribution submission
Accepted contributions will have the opportunity to present their work/ideas in a poster session. Additionally, a few select submissions will have the chance to give a short talk, while others will give a flash talk.
| Time | Session |
|---|---|
| 09:30 – 09:40 | Welcome and introduction by the organizing committee Welcome |
| 09:40 – 10:10 | Invited Talk 1 Invited |
| 10:10 – 10:40 | Invited Talk 2 Invited |
| 10:40 – 11:00 | Contributed flash talks Contributed |
| 11:00 – 11:20 | Coffee break and poster session Break |
| 11:20 – 11:50 | Invited Talk 3 Invited |
| 11:50 – 12:20 | Invited Talk 4 Invited |
| 12:20 – 12:45 | Invited Talk 5 / Panel discussion Invited |
| 12:45 – 12:50 | Farewell and closing remarks Closing |