{"html":"What you'll do
\n\n- Come up with ideas for hard machine learning engineering tasks ("plans") for a long-horizon AI research project with a leading AI lab
\n- Design tasks hard enough to stump frontier models: a strong task is solved in no more than 2 out of 8 attempts
\n- Lean toward inference, evaluation and systems work rather than heavy model training. About half of the tasks run on GPU and half are CPU-only
\n- Explain clearly why each problem is hard for a frontier model
\n- You mainly contribute ideas. You do not need to build or review the tasks yourself
\n
\nMust-haves
\n\n- Current or recent role as an ML research engineer, research scientist or applied scientist
\n- Hands-on experience personally training, fine-tuning or optimizing models in PyTorch or JAX
\n- Deep knowledge of at least one of these areas:
\n- LLMs: pretraining, alignment, evals
\n- GPU inference and performance: vLLM, Triton, CUDA, profiling
\n- Reinforcement learning (RL)
\n- Scientific ML: weather and climate, physics, biology
\n
\n \n- Able to explain why a problem is hard for a frontier model
\n
\nNice-to-haves
\n\n- Experience writing benchmark or evaluation questions (a big plus)
\n- PhD in ML, CS or physics, or an MSc plus strong research output
\n- Publications
\n- Prior Mercor contracts with strong performance
\n
\nNot a fit
\n\n- Software or full-stack engineers whose AI experience is mainly building LangChain or RAG applications
\n- Backgrounds limited to AI trainer, rater or annotator work
\n
"}