ML Research Scientist (AI Task Ideation) - $150-$200/hr

APPLY DIRECTLY HERE: https://t.mercor.com/KHyjW

Role Overview Mercor is partnering with a leading AI lab for a unique, long-horizon research project. We are seeking elite ML Research Engineers, Research Scientists, and Applied Scientists to ideate and design extraordinarily difficult machine learning engineering tasks ("plans") intended to stump frontier AI models.

Crucial Note: In this role, you mainly contribute high-level ideas and task designs. You do NOT need to build, code, or review the tasks yourself.

What You’ll Do

• Design highly complex ML tasks where a frontier model would succeed in no more than 2 out of 8 attempts.
• Focus on inference, evaluation, and systems engineering concepts (approx. half GPU-based, half CPU-only) rather than heavy model training.
• Clearly articulate and document exactly why each problem is uniquely difficult for current frontier models.

Must-Have Qualifications

• Role: Current or recent experience working specifically as an ML Research Engineer, Research Scientist, or Applied Scientist.
• Hands-on ML: Deep, personal experience training, fine-tuning, or optimizing models natively in PyTorch or JAX.
• Deep Domain Expertise in at least ONE of the following:
• LLMs: Pretraining, alignment, or evals.
• GPU Systems: vLLM, Triton, CUDA, or performance profiling.
• Advanced ML: Reinforcement Learning (RL) or Scientific ML (physics, biology, climate).
• Academic/Research Backing: Ability to explain complex model failure modes clearly.

⚠️ Who is NOT a fit:

• Software or Full-Stack engineers whose AI experience is limited to building wrappers, LangChain, or RAG applications.
• Candidates whose background is limited to basic AI training, rating, or data annotation.

Nice-to-Haves: Experience writing benchmark/evaluation questions; a PhD in ML, CS, or Physics; or an MSc with a strong publication record.

Engagement Details & Compensation

• Work Structure: 100% remote, independent contractor position. Highly flexible schedule based entirely on your availability.
• Compensation: Highly competitive flat rate ranging from $150.00 to $200.00 per hour.
• Payments: Processed securely and weekly via Stripe or Wise based on services rendered.

(We consider all qualified applicants without regard to legally protected characteristics and provide reasonable accommodations upon request.)