Ml Scientist - Adversarial Robustness

Mercor; San Francisco (CA); Remote; GBP 103,000 - 124,000; Part time

Job summary

Mercor is seeking an LLM Research Scientist (Pre-training; Computer Vision; Adversarial Robustness) for a remote contract role. You will train image classifiers and generative image models from scratch, and fine-tune open-weight language models.

The role emphasizes robust performance under data and compute constraints and expands capabilities through efficient model compression and troubleshooting. Ideal candidates have 3+ years of ML research, strong experience with PyTorch/JAX/TensorFlow, and

Qualifications

• 3+ years of machine learning research experience (PhD research counts).
• Strong experience with PyTorch, JAX, TensorFlow or similar ML frameworks.
• Degree from a top-100 university, or equivalent research track record through publications/open-source contributions.

Responsibilities

• Train image classifiers and generative image models from scratch.
• Fine-tune open-weight language models.
• Optimize models with limited data, compute, and model-size budgets.
• Enhance model robustness to adversarial inputs and conversations.
• Compress models to meet size and latency constraints without losing accuracy.
• Diagnose and resolve training issues effectively.

Skills

PyTorch

JAX

TensorFlow

ML frameworks

Education

Top-100 university

Tools

—

Job description

About the job

Mercor connects elite creative and technical talent with leading AI research labs. Headquartered in San Francisco, our investors include Benchmark, General Catalyst, Peter Thiel, Adam D'Angelo, Larry Summers, and Jack Dorsey.

Position: LLM Research Scientist (Pre-training & Computer Vision & Adversarial Robustness)
Type: Contract
Compensation: $100–$120/hour
Location: Remote

Role Responsibilities

• Train image classifiers and generative image models from scratch. Fine-tune open-weight language models.
• Optimize models with limited data, compute, and model-size budgets.
• Enhance model robustness to adversarial inputs and conversations.
• Compress models to meet size and latency constraints without losing accuracy.
• Diagnose and resolve training issues effectively.

Qualifications

Must-Have

• 3+ years of machine learning research experience (PhD research counts).
• Strong experience with PyTorch, JAX, TensorFlow, or similar ML frameworks.
• Degree from a top-100 university, experience at a FAANG or comparable AI company, or equivalent research track record through publications or impactful open-source contributions.

Preferred

• Experience in scaling laws and training-efficiency research.
• Knowledge of curriculum learning and data ordering.
• Skills in model evaluation, uncertainty estimation, and model calibration.
• Expertise in data augmentation and synthetic data for robustness.

Resources & Support

• For details about the interview process and platform information, please check: https://talent.docs.mercor.com/welcome
• For any help or support, reach out to: [e-mail]