Supreme Court LLM Data Trainer
Stack
Here are the technologies used in this project:
Research Summary: Key Contributions
Statistical Analysis
- Conducted advanced statistical analyses, including:
- Chi-squared tests.
- Linear regression.
- Confidence intervals.
- Analyzed over 5,000 survey entries with 150+ variables to assess public opinion and its impact on Supreme Court decisions.
Data Visualization
- Developed over 100 lines of code in R, Stata, and Python to create:
- Plots.
- Tables.
- Diagrams.
- Visualized relationships between:
- Public perception and Supreme Court docket fluctuations.
- The influence of SCOTUS decisions on policy.
Database Management
- Structured and managed extensive databases to:
- Analyze survey data.
- Identify trends in SCOTUS case alignments (liberal vs. conservative).
- Examine political beliefs among small businesses.
Research and Publications
- Book Publication:
- Authored and presented a research publication to a board of 10+ professors.
- Conducted meticulous book analysis, proof writing, and academic writing.
- Successfully released the book “Majority Opinions: How an Out-of-Step Supreme Court Can Affect the Rule of Law” with Dr. Neil Malhotra.
- Paper Publication:
- Collaborated on a paper with Dr. Malhotra: “The Politics of Small Business Owners.”
AI Development
- Fine-tuned an AI model leveraging Large Language Models (LLMs) to:
- Predict Supreme Court decisions.
- Process over 1,500 files, including 800 court documents and personal appeals.
- Classify cases based on specific justices.
- Engineered Python-based SCOTUS case file data loaders for Google Cloud processors, optimizing pre-trained LLM models.
Future Directions
- Explore deeper intersections between public opinion and judicial decisions.
- Expand AI models to incorporate real-time data for predictive analysis.
- Continue academic publishing to further the understanding of SCOTUS’s influence on policy and public sentiment.