
Quant Development - AI/ML
- Singapore
- Permanent
- Full-time
- Advanced model architectures, including sparse activation mechanisms and dynamic routing algorithms
- Efficient training and inference methods for large-scale models
- Innovations in attention mechanisms, tokenization algorithms, and embedding optimization
- Developing novel approaches to mixture of experts, multi-head attention, and tokenization
- Analyzing the impact of tokenization on model performance and designing unified frameworks for multilingual models
- Optimizing embedding compression and semantic space optimization
- A Master's or PhD in Computer Science, AI, Mathematics, or a related field
- Research experience in relevant areas, with published papers or in-depth projects
- Proficiency in frameworks like PyTorch/JAX and experience with large model training and fine-tuning
- A deep understanding of Transformer architectures and experience with related source code
- Familiarity with distributed training and memory optimization is desirable