AI Data Labeller (RLHF projects)
Worked on RLHF-focused AI tasks including comparing model responses and performing quality checks. Authored and refined prompts intended to guide models toward better, clearer answers. Ensured consistency and correctness of outputs for downstream use. • Compared model responses to assess quality and adherence • Performed quality review and error flagging • Wrote prompts/instructions to improve answer quality • Supported instruction following and safer output behavior