M.Tech in Frontier AI Systems
2 years · 60 places · School of Computing
- Foundation models: pre-training, scaling laws, efficient fine-tuning, distillation.
- Agentic AI: multi-agent orchestration, tool use, planning and long-horizon reasoning.
- AI safety and alignment: interpretability, red-teaming, evaluation science, governance.
- Multimodal and embodied AI: vision-language models, robotics, simulation-to-real transfer.
- Edge and sovereign AI: on-device inference, Indic-language models, low-resource computing.
- Capstone: build and deploy a model on the Aryabhata-X cluster (4,096 GPUs).