Member of Technical Staff, Pretraining evaluations

As a Member of Technical Staff in Pretraining Evaluations at Cohere, you will be instrumental in improving the evaluation processes for large language models (LLMs). Your responsibilities will include enhancing evaluation benchmarks, implementing new measurement tasks, and improving the statistical framework to reduce noise and increase the signal-to-noise ratio in evaluations. The role requires a strong background in statistics, data science, and software engineering, alongside proficient programming skills in languages such as Python and familiarity with machine learning frameworks. Collaboration with cross-functional teams and effective communication of findings will be essential.