AI Trainer — Mathematical Reasoning
Text
Designed and authored adversarial mathematics and reasoning challenges to expose model failure modes in mathematical domains. Reviewed and evaluated AI-generated solutions for accuracy, stepwise reasoning, and logical validity. Authored structured feedback for use in refining AI model training data. • Created algebra, calculus, geometry, trigonometry, and math word problem tasks.
• Performed quality assurance and reasoning validation of AI outputs.
• Developed annotation and test-set guidelines for mathematical reasoning.
• Utilized feedback to improve AI model performance and address failure cases.
Designed and authored adversarial mathematics and reasoning challenges to expose model failure modes in mathematical domains. Reviewed and evaluated AI-generated solutions for accuracy, stepwise reasoning, and logical validity. Authored structured feedback for use in refining AI model training data. • Created algebra, calculus, geometry, trigonometry, and math word problem tasks. • Performed quality assurance and reasoning validation of AI outputs. • Developed annotation and test-set guidelines for mathematical reasoning. • Utilized feedback to improve AI model performance and address failure cases.
2024 - Present