Machine Learning Research Scientist

This is a fully remote position, open to applicants in New York.

📋 Description

• Train image classifiers and generative image models from the ground up.

• Fine-tune and perform post-training on open-weight language models.

• Design and implement empirical experiments in machine learning.

• Identify and resolve issues related to optimization, convergence, data quality, and training stability.

• Create methods that enhance performance while working within constraints of data, computational resources, or model size.

• Train computer vision and generative models, such as diffusion models, GANs, VAEs, and flow-based architectures.

• Assess generative models using evaluation metrics like FID.

• Enhance sample quality while managing training costs and parameter counts.

• Develop models that are resilient to adversarial inputs and assess their robust accuracy.

• Implement adversarial training, quantization, pruning, knowledge distillation, and other model-compression strategies.

• Carry out supervised fine-tuning and preference optimization utilizing techniques like DPO, RLHF, or RLAIF.

• Create training datasets through methods such as synthetic generation, weak supervision, noisy supervision, or rejection sampling.

• Enhance multi-turn conversational capabilities, calibrated confidence, and responsiveness to corrections.

• Train multilingual and low-resource language models, along with developing tokenization and sampling strategies.

• Investigate cross-lingual transfer and boost performance in data-scarce language environments.

• Collaborate with seasoned AI researchers on complex technical endeavors.

• Conduct research in areas such as robustness, efficiency, generative modeling, post-training, multilingual modeling, and evaluation.

⛳️ Requirements

• A minimum of 3 years of experience in machine learning research, including qualifying PhD research.

• Practical experience in training deep learning models from start to finish.

• Strong expertise in machine learning frameworks such as PyTorch, JAX, or TensorFlow.

• In-depth knowledge in at least one pertinent field: adversarial robustness, computer vision, generative modeling, LLM post-training, or multilingual pre-training.

• Experience in designing and conducting rigorous empirical experiments.

• Solid understanding of optimization, model evaluation, and experimental techniques.

• Capability to troubleshoot complex issues related to model training and performance.

• Excellent technical writing and research communication abilities.

• A degree in computer science, machine learning, artificial intelligence, mathematics, statistics, engineering, or a related technical discipline is highly relevant.

• PhD research in machine learning or a closely related area can contribute to fulfilling the professional experience requirement.

• Equivalent research proficiency may also be demonstrated through significant industry experience, publications, or impactful open-source contributions.

• Desirable experience includes knowledge of scaling laws, training efficiency, curriculum learning, benchmarks, contamination detection, uncertainty estimation, synthetic data, or data augmentation.

• Publications in reputable machine learning or AI conferences are a plus.

• Experience with a leading AI, technology, or research organization is advantageous.

• Substantial contributions to open-source machine learning projects are beneficial.

🏝️ Benefits

• Flexible scheduling options.

• Competitive hourly pay ranging from $95 to $115.

• Weekly payments via Stripe or Wise.

• Flexibility in project-based assignments.

• Projects can be extended, shortened, or modified based on scope and performance.