CoMI-IRL: Contrastive Multi-Intention Inverse Reinforcement Learning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866911431021035520 |
|---|---|
| author | Mone, Antonio Oliehoek, Frans A. Siebert, Luciano Cavalcante |
| author_facet | Mone, Antonio Oliehoek, Frans A. Siebert, Luciano Cavalcante |
| contents | Inverse Reinforcement Learning (IRL) seeks to infer reward functions from expert demonstrations. When demonstrations originate from multiple experts with different intentions, the problem is known as Multi-Intention IRL (MI-IRL). Recent deep generative MI-IRL approaches couple behavior clustering and reward learning, but typically require prior knowledge of the number of true behavioral modes $K^*$. This reliance on expert knowledge limits their adaptability to new behaviors, and only enables analysis related to the learned rewards, and not across the behavior modes used to train them. We propose Contrastive Multi-Intention IRL (CoMI-IRL), a transformer-based unsupervised framework that decouples behavior representation and clustering from downstream reward learning. Our experiments show that CoMI-IRL outperforms existing approaches without a priori knowledge of $K^*$ or labels, while allowing for visual interpretation of behavior relationships and adaptation to unseen behavior without full retraining. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_07496 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | CoMI-IRL: Contrastive Multi-Intention Inverse Reinforcement Learning Mone, Antonio Oliehoek, Frans A. Siebert, Luciano Cavalcante Machine Learning Inverse Reinforcement Learning (IRL) seeks to infer reward functions from expert demonstrations. When demonstrations originate from multiple experts with different intentions, the problem is known as Multi-Intention IRL (MI-IRL). Recent deep generative MI-IRL approaches couple behavior clustering and reward learning, but typically require prior knowledge of the number of true behavioral modes $K^*$. This reliance on expert knowledge limits their adaptability to new behaviors, and only enables analysis related to the learned rewards, and not across the behavior modes used to train them. We propose Contrastive Multi-Intention IRL (CoMI-IRL), a transformer-based unsupervised framework that decouples behavior representation and clustering from downstream reward learning. Our experiments show that CoMI-IRL outperforms existing approaches without a priori knowledge of $K^*$ or labels, while allowing for visual interpretation of behavior relationships and adaptation to unseen behavior without full retraining. |
| title | CoMI-IRL: Contrastive Multi-Intention Inverse Reinforcement Learning |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2602.07496 |