Analista de Qualidade e Treinamento de IA (Freelancer)
OtherTextRLHF
I was responsible for the training and evaluation of generative AI models through Reinforcement Learning from Human Feedback. My work included refining AI output for accuracy and safety, as well as comparative performance testing among large language models. Additionally, I led the labeling and classification of multiple data types, focusing on reducing bias and hallucinations in model outputs.
• Conducted RLHF-based training for generative language models.
• Performed A/B evaluation and ranking of LLM responses.
• Labeled and classified multimodal data—text, image, and audio.
• Audited language models for bias mitigation and factual accuracy.
I was responsible for the training and evaluation of generative AI models through Reinforcement Learning from Human Feedback. My work included refining AI output for accuracy and safety, as well as comparative performance testing among large language models. Additionally, I led the labeling and classification of multiple data types, focusing on reducing bias and hallucinations in model outputs. • Conducted RLHF-based training for generative language models. • Performed A/B evaluation and ranking of LLM responses. • Labeled and classified multimodal data—text, image, and audio. • Audited language models for bias mitigation and factual accuracy.
2025 - 2025