Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2506.01623 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918126283653120 |
|---|---|
| author | Palattuparambil, Ajsal Shereef Karimpanal, Thommen George Rana, Santu |
| author_facet | Palattuparambil, Ajsal Shereef Karimpanal, Thommen George Rana, Santu |
| contents | Humans excel at analogical reasoning - applying knowledge from one task to a related one with minimal relearning. In contrast, reinforcement learning (RL) agents typically require extensive retraining even when new tasks share structural similarities with previously learned ones. In this work, we propose MAGIK, a novel framework that enables RL agents to transfer knowledge to analogous tasks without interacting with the target environment. Our approach leverages an imagination mechanism to map entities in the target task to their analogues in the source domain, allowing the agent to reuse its original policy. Experiments on custom MiniGrid and MuJoCo tasks show that MAGIK achieves effective zero-shot transfer using only a small number of human-labelled examples. We compare our approach to related baselines and highlight how it offers a novel and effective mechanism for knowledge transfer via imagination-based analogy mapping. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_01623 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MAGIK: Mapping to Analogous Goals via Imagination-enabled Knowledge Transfer Palattuparambil, Ajsal Shereef Karimpanal, Thommen George Rana, Santu Artificial Intelligence Machine Learning Humans excel at analogical reasoning - applying knowledge from one task to a related one with minimal relearning. In contrast, reinforcement learning (RL) agents typically require extensive retraining even when new tasks share structural similarities with previously learned ones. In this work, we propose MAGIK, a novel framework that enables RL agents to transfer knowledge to analogous tasks without interacting with the target environment. Our approach leverages an imagination mechanism to map entities in the target task to their analogues in the source domain, allowing the agent to reuse its original policy. Experiments on custom MiniGrid and MuJoCo tasks show that MAGIK achieves effective zero-shot transfer using only a small number of human-labelled examples. We compare our approach to related baselines and highlight how it offers a novel and effective mechanism for knowledge transfer via imagination-based analogy mapping. |
| title | MAGIK: Mapping to Analogous Goals via Imagination-enabled Knowledge Transfer |
| topic | Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2506.01623 |