Sandrine L.

I am a native French speaker with a BA in English specializing in literature, linguistics, and translation. My areas of expertise include: • Training large language models (LLMs) to enhance their accuracy and linguistic capabilities in French and English. • Translating between English and French (general, literary, and specialized content). • Proofreading AI-translated material to ensure accuracy, fluency, and cultural appropriateness. AI Training Experience: -Adversarial Prompt Engineering: designing complex prompt sequences in order to identify failure modes and edge cases in large language models, contributing to improved model reliability and safety. -Instruction Following Evaluation: crafting complex, multi-instruction prompts to assess models' ability to accurately interpret and execute layered directives. -Rubric-Based Quality Assessment: applying standardized evaluation frameworks to assess model performance against quality benchmarks, supporting continuous model improvement. -Red Teaming Initiatives: designing prompts to identify potential weaknesses, biases and safety issues in language models, helping improve model safeguards and policies. -Audio Labeling & Annotation: performing precise audio content annotation to support multimodal model training and improve speech recognition capabilities.