AI Tutor/Trainer
As an AI Tutor/Trainer at Mindrift (Powered by Toloka), I specialized in optimizing Large Language Models by generating high-quality datasets and providing expert feedback. My work focused on evaluating and ranking model outputs using advanced Reinforcement Learning from Human Feedback (RLHF) tasks. I continually identified and corrected hallucinations, biases, and logical inaccuracies to ensure model reliability. • Conducted detailed audits of AI-generated text content for relevance and accuracy. • Provided fine-tuned feedback and corrections for LLM optimization. • Executed RLHF evaluation and ranking tasks to improve conversational AI models. • Collaborated with the Mindrift team to meet quality and project goals.