AI Trainer (LLM / RLHF Annotation)
Responsible for ranking LLM model outputs, evaluating quality, and providing feedback for reinforcement learning with human feedback (RLHF) workflows. Assessed and annotated prompt-response pairs, reviewed AI-generated outputs for hallucinations, and contributed to output quality improvement. Utilized various RLHF tasks including preference ranking, response rating, and red teaming for ethical/robust model performance. • Applied RLHF concepts to improve language model alignment. • Rated AI responses using output evaluation frameworks. • Provided annotation and quality review for hallucination detection. • Supported continuous model improvement through structured feedback.