ML Scientist: Adversarial Robustness & Generative Vision

Mercor; Chicago (IL); On-site; USD 120,000 - 180,000; Full time

Job summary

Mercor seeks experienced machine learning researchers to train and improve deep learning models end-to-end, across vision and language. You will tackle well-scoped empirical ML research problems and work on robustness, efficiency, and performance.

Responsibilities include training classifiers and generative models, reducing data and compute footprints, and ensuring robust behavior under adversarial conditions. Collaboration with top researchers is expected.

Qualifications

• 3+ years of machine learning research experience (PhD counts toward this).
• Strong experience with major ML frameworks and research practices.
• Publication record or impactful contributions are valued.

Responsibilities

• Train image classifiers and generative models end-to-end across vision and language.
• Improve robustness against adversarial inputs and conversations.
• Compress models to meet size and latency limits without sacrificing accuracy.
• Diagnose and resolve training issues and iterate on improvements.
• Collaborate with researchers on high-impact projects.

Skills

Adversarial Robustness

Efficient CV

Generative Models

LLM Fine-tuning

Multilingual Training

Data Augmentation

Alignment Fine-tuning

Uncertainty Estimation

Curriculum Learning

Model Evaluation

Education

PhD or Equivalent

Tools

PyTorch

JAX

TensorFlow

Job description

Mercor seeks experienced machine learning researchers to train and improve deep learning models end-to-end, across vision and language. You will tackle well-scoped empirical ML research problems and work on robustness, efficiency, and performance.

Responsibilities include training classifiers and generative models, reducing data and compute footprints, and ensuring robust behavior under adversarial conditions. Collaboration with top researchers is expected.