AI Optimization & LLM Evaluation

Experience evaluating and improving Large Language Models (LLMs) at Outlier AI.

Beyond traditional physics research, I have hands-on experience with cutting-edge AI and machine learning technologies, particularly in the domain of Large Language Models (LLMs).

Work at Outlier AI

As part of the Outlier AI team, I contributed to:

  • Model Evaluation: Systematic assessment of LLM outputs across various domains
  • Quality Assurance: Rigorous testing of model responses for accuracy, coherence, and safety
  • Performance Optimization: Identifying patterns in model behavior to guide improvements
  • Domain Expertise: Applying physics and mathematical knowledge to evaluate technical content

Key Skills

This experience has strengthened my abilities in:

  • Critical evaluation of AI-generated content
  • Understanding of modern NLP architectures and their limitations
  • Interdisciplinary problem-solving combining domain expertise with AI/ML
  • Quality metrics and evaluation frameworks

Bridging Physics and AI

The intersection of high-energy physics and artificial intelligence is particularly exciting. Modern HEP analyses increasingly rely on machine learning for:

  • Event classification and particle identification
  • Anomaly detection in detector systems
  • Optimization of trigger algorithms
  • Fast simulation techniques

My experience with both traditional physics analysis and modern AI systems positions me well to contribute to this evolving landscape.

Technical Proficiency

  • Programming: Python, C++, ROOT
  • ML Frameworks: PyTorch, TensorFlow, scikit-learn
  • HEP Tools: Coffea, Uproot, Awkward Array
  • Data Analysis: Pandas, NumPy, SciPy
  • Version Control: Git, GitHub