Job Overview

• 6-month full-time research internship (extension/full-time conversion possible)
• Work on frontier AI research in alignment, interpretability & safety
• Contribute to large-scale foundation model research
• Collaborate in a flat, high-performance research environment
• Exposure to cutting-edge Superintelligence research methodologies
• Opportunity to co-author research papers & whitepapers

Role

• Develop and enhance open-source Python libraries for alignment & explainability
• Conduct experiments in:
• Explainability (XAI techniques like SHAP, LRP, Grad-CAM, etc.)
• Mechanistic Interpretability (activation patching, circuit analysis)
• Uncertainty estimation & robustness benchmarking
• Reinforcement Learning & RL fine-tuning
• Tabular Foundational Models (Orion, TabTune)
• Implement and benchmark model optimization techniques
• Author experiment code & research documentation
• Collaborate with research & engineering teams
• Contribute to conference papers & technical reports

Salary

• Competitive Stipend
• Potential Full-Time Offer based on performance
• Access to high-end compute (GPUs & cloud credits)

Location

• Not explicitly mentioned (Likely Remote / Research-based setup)

Work Timing

• Full-Time Internship
• 6 Months Commitment
• Rolling Start Date
• Standard Research Team Schedule (as per project needs)

Eligibility

• Strong Python programming skills (modular & production-level code)
• Deep understanding of ML/DL fundamentals
• Hands-on experience with PyTorch
• Strong knowledge of Transformer architectures (GPT, BERT, LLaMA, T5, etc.)
• Experience with Git & collaborative workflows
• Excellent research & communication skills