Hark, an AI company in San Jose, seeks a Member of Technical Staff for Pretraining Evaluations to build evaluation harnesses spanning text, code, math, and multimodal content. You will design benchmarks, track held-out loss, and compare training runs to guide recipe choices and checkpoint selection.
The role emphasizes robust Python tooling, statistical rigor, and collaboration with researchers to extend evaluations to new modalities and datasets.
#J-18808-Ljbffr
The following opening is for a Staff AI Evaluation Engineer – Pretraining with Hark.
This role, Staff AI Evaluation Engineer – Pretraining at Hark, could be your next move.
Are you ready to take on the Staff AI Evaluation Engineer – Pretraining role at Hark?
We would love to welcome a new Staff AI Evaluation Engineer – Pretraining to our team in San Jose, CA, United States.