AI & ML Post-Graduate Research Mentorship

Research Mentorship for MSc & PhD Graduate Students

Remote | Worldwide

EnergyAI Berlin is looking for motivated MSc and PhD students to join our mentorship program focused on generative AI research.

What you'll get

• Work with like-minded individuals on breakthrough GenAI research
• Regular meetings and hands-on guidance throughout all stages of your research
• Compute access

Research topics

Deep learning research track

• Architectural design & mechanistic interpretability on very small models
• Training and evaluation of very small language models (5100M parameters)
• LLM harness R&D (esp. coding & Recursive LM) SWE experience preferred

Domain-specific track (ML/applied math + domain expertise)

• Biologically-inspired network component design/evaluation backgrounds in neuroscience, evolutionary systems, or network/consensus algorithms
• Cultural alignment evaluation of LLMs backgrounds in sociology, anthropology, linguistics, or history

Requirements

• Deep experience with PyTorch/JAX (ML arch/eng topics) or NumPy/Pandas (data-centric topics)
• Solid statistics background; linear algebra for non-data-centric topics
• Comfortable with Linux and remote work
• Basic front-end dev familiarity is a plus (even with assistant help)

About EnergyAI Berlin

EnergyAI is a Berlin-based AI R&D company building super-specialized language models for the energy sector. GridOS is our integrated solution for grid operators, energy utilities, and e-mobility providers tackling challenges in grid management, power flow analysis, and workflow automation. Our research arm explores small, domain-specialized language models as an alternative to general-purpose LLMs.

Careers

https://energyai.berlin/career

Interested

• Submit your application at:https://forms.gle/MberzSKLs85qqw4YA
• For inquiries, email us at: [Confidential Information] with the subject line Research Mentorship Application [Your Name]

#GenAI #Research #Mentorship #PhD #MSc #MachineLearning #DeepLearning #NLP #RemoteWork #OpenScience