Kathleen Torey

Key Skills

Software

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Text

Top Label Types

Question Answering

Text Generation

RLHF

Fine Tuning

Evaluation Rating

Freelancer Overview

I am a mathematics professional with extensive experience in AI training data creation, data labeling, and rubric-based evaluation for complex mathematical domains. My background includes designing and annotating university-level mathematics prompts, authoring step-by-step solutions, and developing structured scoring rubrics to assess reasoning depth and formal rigor. I have a strong command of LaTeX and am skilled at tagging datasets with detailed metadata, such as topic, difficulty, and error classification, to enhance the quality and reliability of AI models. My work emphasizes logical consistency, precision, and the ability to follow intricate technical guidelines, making me highly effective in projects that require careful data annotation and evaluation for AI training. With advanced English proficiency and a proven track record in both academic tutoring and AI response assessment, I am dedicated to improving model accuracy and interpretability through meticulous data preparation.

Labeling Experience

University-Level Mathematics Reasoning Dataset – RLHF & Evaluation Project

Scale AITextQuestion AnsweringText Generation

Contributed to the development and evaluation of a large-scale university-level mathematics reasoning dataset used for training and fine-tuning large language models (LLMs). Designed challenging prompts across algebra, calculus, probability and statistics, linear algebra, discrete mathematics, and optimization, ensuring coverage of both computational and proof-based reasoning tasks. Authored complete, step-by-step solutions emphasizing logical rigor, formal justification, and mathematical clarity. Developed structured scoring rubrics to assess AI-generated responses based on correctness, completeness, reasoning depth, and adherence to formal proof standards. Performed detailed evaluation and rating of model outputs within RLHF frameworks, identifying logical inconsistencies, computational errors, unsupported assumptions, and gaps in multi-step reasoning. Provided corrected solutions with concise rationales to improve model alignment and reliability.

Contributed to the development and evaluation of a large-scale university-level mathematics reasoning dataset used for training and fine-tuning large language models (LLMs). Designed challenging prompts across algebra, calculus, probability and statistics, linear algebra, discrete mathematics, and optimization, ensuring coverage of both computational and proof-based reasoning tasks. Authored complete, step-by-step solutions emphasizing logical rigor, formal justification, and mathematical clarity. Developed structured scoring rubrics to assess AI-generated responses based on correctness, completeness, reasoning depth, and adherence to formal proof standards. Performed detailed evaluation and rating of model outputs within RLHF frameworks, identifying logical inconsistencies, computational errors, unsupported assumptions, and gaps in multi-step reasoning. Provided corrected solutions with concise rationales to improve model alignment and reliability.

2023 - 2025

Education

Work History