4 months ago
Paris, France +3 moreSenior
Responsibilities
- Research, design, and train large-scale foundational models for the physical world.
- Train and fine-tune Vision-Language Models on custom datasets.
- Own the full model lifecycle from data curation and training through evaluation.
- Implement research ideas rapidly from current ArXiv papers.
- Develop custom layers, loss functions, and distributed training loops.
- Build data pipelines for large-scale datasets.
Requirements
- MSc or PhD in Machine Learning, Computer Science, or Mathematics with a focus on Deep Learning.
- Theoretical understanding and practical experience training foundational models such as LLMs, VLMs, or other large-scale multimodal models from scratch or through fine-tuning.
- Strong software engineering skills in Python and fluency with PyTorch or JAX.
- Experience writing custom layers, loss functions, and distributed training loops.
- Clear communication skills and the ability to contribute to an internal research culture.
- Preferred: publications at top-tier conferences including NeurIPS, ICML, ICLR, CVPR, or ACL on attention mechanisms, efficient transformers, or multimodal learning.
- Preferred: experience training models on large-scale GPU clusters.
- Preferred: experience with data curation, cleaning, pruning, and robust pipelines for large-scale datasets.
Benefits
- Competitive salary and VSOP options
- Relocation support of up to €2,500 and four weeks of temporary accommodation
- €500/£450 yearly learning allowance
- Gym membership and mental health support through Nilo.health
- Regular company events and monthly social allowances
- Enhanced parental leave of 22 weeks for primary caregivers and six weeks for secondary caregivers
- Five days of paid family emergency leave, 100% remote work during pregnancy, and phased return to work
Categories
AI ResearchML Engineering
About Helsing
Helsing was founded to provide Artificial Intelligence to protect our democracies. Helsing works with partners in governments and across industry. Together, we develop systems that leverage AI and autonomy to provide autonomous mass that is sovereign, scalable and affordable. We’re looking for people with their heart in the right place, who share our conviction that liberal democratic values are worth protecting, for ourselves and for future generations.