LLM Research Scientist (Pre-training & Computer Vision & Adversarial Robustness)

Science & AI Safety • Independent Contractor • Remote • Weekly Payouts

We're looking for experienced machine learning researchers with hands-on experience training and improving deep learning models end-to-end, across vision and language. You'll work on well-scoped empirical open-ended ML research problems.

Responsibilities

• Train image classifiers and generative image models from scratch, and fine-tune open-weight language models.
• Get the most out of limited data, compute, and model-size budgets.
• Make models robust, to adversarial inputs and to adversarial conversations.
• Compress models to meet hard size and latency constraints without sacrificing accuracy.
• Diagnose and resolve training issues.

Requirements

We are looking for candidates with strong expertise in one or more of the following areas:

Adversarial Robustness

Experience with

• Adversarial training of image classifiers (e.g. PGD-based training, TRADES).
• Evaluating robust accuracy under standard threat models (e.g. L∞ attacks, AutoAttack) and avoiding gradient-masking pitfalls.
• Managing the robustness, accuracy trade-off and robust overfitting.

Efficient Computer Vision

Experience with

• Training image classifiers end-to-end, especially for fine-grained recognition (many visually similar classes, few examples per class).
• Model compression: quantization, pruning, and knowledge distillation from large teachers into small students.
• Deploying models under hard size or latency budgets (on-device, edge, or embedded settings).

Generative Image Modeling

Experience with

• Training image generative models from scratch: diffusion models, GANs, VAEs, or flow-based models.
• Iterating against sample-quality metrics such as FID.
• Training-efficiency tricks that produce good generators quickly and at small parameter counts.

LLM Post-Training & Behavioral Robustness

Hands-on experience with one or more of

• Supervised fine-tuning and preference optimisation (DPO, RLHF, RLAIF) of open-weight language models, including building your own datasets via synthetic generation, noisy or weak supervision, and rejection sampling.
• Shaping conversational behaviour over multiple turns: resistance to persuasion and sycophancy, calibrated confidence, and knowing when to accept corrections.
• Alignment-style fine-tuning that changes a specific behaviour while preserving general capability.

Multilingual Pre-training

Experience with

• Training multilingual or low-resource-language models from scratch.
• Tokenizer design across scripts and typologically diverse languages.
• Balancing highly unequal per-language data (sampling temperatures, cross-lingual transfer) in data-constrained regimes.

Additional Areas of Interest

Experience in any of the following is a plus

• Scaling laws and training-efficiency research.
• Curriculum learning and data ordering.
• Model evaluation: benchmark construction, contamination control, statistically sound comparisons.
• Uncertainty estimation and model calibration.
• Data augmentation and synthetic data for robustness.

General Qualifications

• 3+ years of machine learning research experience (PhD research counts toward this requirement).
• 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 an equivalent research track record through publications or impactful open-source contributions.

Why Join

• Work on cutting-edge machine learning research.
• Collaborate with leading AI researchers on challenging, high-impact projects.
• Flexible, project-based work with competitive compensation.

Details

• Pay: $100 to $120 per hour (hourly)
• Commitment: 40 hours per week • remote
• Eligible locations: United States
• Platform: Mercor (weekly payouts via Stripe or Wise)

Submit your application via the link below. Qualified candidates move through Mercor's short AI interview and selection process. Projects can be extended, shortened, or concluded early depending on needs and performance.