Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2511.20610 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915637583937536 |
|---|---|
| author | Merten, Gaspard Sakr, Mahmoud Dejaegere, Gilles |
| author_facet | Merten, Gaspard Sakr, Mahmoud Dejaegere, Gilles |
| contents | Foundation models are transformative in artificial intelligence, but building them from scratch, especially for mobility trajectories, is not yet clear or documented. This tutorial bridges this gap by demonstrating the steps and code of a minimal implementation of a trajectory-focused foundation model starting from GPT-2. Through a concise, step-by-step, code-driven process, we demonstrate adapting GPT-2 for spatiotemporal data. We then review and compare representative trajectory foundation models, such as TrajFM and TrajGPT, highlighting their architectural innovations and differences. Additionally, we introduce complementary techniques from related domains, like TimesFM's patching approach. Targeted at researchers and practitioners, this tutorial aims to explain the concepts and terminology of foundation models, at the implementation level. We find it timely and indispensable to create this educational material in order to support the SIGSPATIAL community in building and evaluating mobility foundation models, enhancing both research clarity and peer-review effectiveness in mobility AI. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_20610 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Building a Foundation Model for Trajectory from Scratch Merten, Gaspard Sakr, Mahmoud Dejaegere, Gilles Artificial Intelligence Foundation models are transformative in artificial intelligence, but building them from scratch, especially for mobility trajectories, is not yet clear or documented. This tutorial bridges this gap by demonstrating the steps and code of a minimal implementation of a trajectory-focused foundation model starting from GPT-2. Through a concise, step-by-step, code-driven process, we demonstrate adapting GPT-2 for spatiotemporal data. We then review and compare representative trajectory foundation models, such as TrajFM and TrajGPT, highlighting their architectural innovations and differences. Additionally, we introduce complementary techniques from related domains, like TimesFM's patching approach. Targeted at researchers and practitioners, this tutorial aims to explain the concepts and terminology of foundation models, at the implementation level. We find it timely and indispensable to create this educational material in order to support the SIGSPATIAL community in building and evaluating mobility foundation models, enhancing both research clarity and peer-review effectiveness in mobility AI. |
| title | Building a Foundation Model for Trajectory from Scratch |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2511.20610 |