Bilevel Planning With Learned Symbolic Abstractions From Interaction Data
IEEE Robotics and Automation Letters, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Basım Tarihi: 2026
- Doi Numarası: 10.1109/lra.2026.3732904
- Dergi Adı: IEEE Robotics and Automation Letters
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus
- Anahtar Kelimeler: Bilevel Planning, Developmental Robotics, Learning Categories and Concepts, Learning from Experience
- Orta Doğu Teknik Üniversitesi Adresli: Hayır
Özet
Intelligent agents must reason over both continuous dynamics and discrete representations to generate effective plans in complex environments. Previous studies have shown that symbolic abstractions can emerge from neural effect predictors trained on a robot's unsupervised exploration. However, these methods rely on deterministic symbolic domains, lack mechanisms to verify the generated symbolic plans, and operate only at the abstract level, often failing to capture the continuous dynamics of the environment. To overcome these limitations, we propose a bilevel neuro-symbolic framework in which learned probabilistic symbolic rules generate candidate plans rapidly at the high level, and learned continuous effect models verify these plans and perform forward search when necessary at the low level. Our experiments on multi-object manipulation tasks demonstrate that the proposed bilevel planning method outperforms symbolic baselines, achieves the planning success of continuous forward search at a fraction of its computational cost, with a verification mechanism that reliably validates symbolic plans and guides the transition between planning levels.