DriveCode: Domain Specific Numerical Encoding for LLM-Based Autonomous Driving
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910061636354048 |
|---|---|
| author | Wang, Zhiye Jiang, Yanbo Zhou, Rui Zhang, Bo Zhang, Fang Xu, Zhenhua Zhang, Yaqin Wang, Jianqiang |
| author_facet | Wang, Zhiye Jiang, Yanbo Zhou, Rui Zhang, Bo Zhang, Fang Xu, Zhenhua Zhang, Yaqin Wang, Jianqiang |
| contents | Large language models (LLMs) have shown great promise for autonomous driving. However, discretizing numbers into tokens limits precise numerical reasoning, fails to reflect the positional significance of digits in the training objective, and makes it difficult to achieve both decoding efficiency and numerical precision. These limitations affect both the processing of sensor measurements and the generation of precise control commands, creating a fundamental barrier for deploying LLM-based autonomous driving systems. In this paper, we introduce DriveCode, a novel numerical encoding method that represents numbers as dedicated embeddings rather than discrete text tokens. DriveCode employs a number projector to map numbers into the language model's hidden space, enabling seamless integration with visual and textual features in a unified multimodal sequence. Evaluated on OmniDrive, DriveGPT4, and DriveGPT4-V2 datasets, DriveCode demonstrates superior performance in trajectory prediction and control signal generation, confirming its effectiveness for LLM-based autonomous driving systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_00919 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DriveCode: Domain Specific Numerical Encoding for LLM-Based Autonomous Driving Wang, Zhiye Jiang, Yanbo Zhou, Rui Zhang, Bo Zhang, Fang Xu, Zhenhua Zhang, Yaqin Wang, Jianqiang Computer Vision and Pattern Recognition Robotics Large language models (LLMs) have shown great promise for autonomous driving. However, discretizing numbers into tokens limits precise numerical reasoning, fails to reflect the positional significance of digits in the training objective, and makes it difficult to achieve both decoding efficiency and numerical precision. These limitations affect both the processing of sensor measurements and the generation of precise control commands, creating a fundamental barrier for deploying LLM-based autonomous driving systems. In this paper, we introduce DriveCode, a novel numerical encoding method that represents numbers as dedicated embeddings rather than discrete text tokens. DriveCode employs a number projector to map numbers into the language model's hidden space, enabling seamless integration with visual and textual features in a unified multimodal sequence. Evaluated on OmniDrive, DriveGPT4, and DriveGPT4-V2 datasets, DriveCode demonstrates superior performance in trajectory prediction and control signal generation, confirming its effectiveness for LLM-based autonomous driving systems. |
| title | DriveCode: Domain Specific Numerical Encoding for LLM-Based Autonomous Driving |
| topic | Computer Vision and Pattern Recognition Robotics |
| url | https://arxiv.org/abs/2603.00919 |