4 months ago
Base Salary
$175k - $296k/yr
Responsibilities
- Design and implement large-scale multimodal vision-language-action transformer architectures for end-to-end autonomous driving.
- Develop pretraining and fine-tuning strategies using labeled and unlabeled fleet data, including images, video, LiDAR, CAN bus data, maps, and human driving behaviors.
- Research cross-modal alignment, visual grounding, temporal reasoning, policy distillation, imitation learning, and reinforcement learning methods.
- Scale model training across thousands of GPUs using distributed training frameworks such as FSDP and DDP.
- Conduct ablation studies, evaluation, and visualization of model behavior across perception, reasoning, and planning tasks.
- Optimize models for onboard deployment through quantization, export, and latency-accuracy trade-offs.
- Collaborate with infrastructure, perception, and planning engineers to deliver production-ready models.
Requirements
- Master’s degree or higher in Computer Science, Electrical/Computer Engineering, or a related field, with 3+ years of deep learning research or productization experience.
- Strong proficiency in PyTorch and modern transformer-based model design.
- Experience with large-scale pretraining or multimodal modeling involving vision, language, or planning.
- Deep understanding of representation learning, temporal modeling, and self-supervised or reinforcement learning techniques.
- Familiarity with distributed training using DDP and FSDP and with large-batch optimization.
- Preferred qualifications include a PhD in CS, CE, EE, or a related field with 1+ years of relevant industry experience.
- Preferred qualifications include publications at CVPR, ICCV, NeurIPS, ICLR, ICML, or ECCV.
- Preferred experience includes foundation models, end-to-end driving models, LLM/VLM architectures, ViT, Flamingo, BEVFormer, RT-2, GRPO-style policies, RLHF, DPO, trajectory prediction, or policy learning for control tasks.
- Ability to collaborate cross-functionally with infrastructure, perception, and planning teams.
Benefits
- Collaborative, research-driven environment with access to massive real-world data and industry-scale compute.
- Opportunity to work with top-tier researchers and engineers on foundation models for autonomous driving.
- Direct impact on next-generation intelligent mobility and autonomous-driving systems.
- Competitive compensation package.
- Snacks, lunches, dinners, and fun activities.
Tech Stack
Categories
AI ResearchML Engineering
About XPENG
XPENG designs, manufactures, and sells smart electric vehicles for consumers, integrating in-house advanced driver-assistance systems and connected in-car software. Founded in 2014 and headquartered in Guangzhou, it operates plants in Zhaoqing and Guangzhou, maintains a European HQ in Amsterdam, and develops eVTOL aircraft via XPENG AEROHT.