LLM Research Scientist – Pre-training, Computer Vision, Adversarial Robustness

This is a fully remote position, open to applicants in United States.

📋 Description

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

• Fine-tune language models that are open-weight.

• Address empirical, open-ended research challenges in machine learning.

• Optimize models while adhering to constraints on data, computational resources, and model size.

• Enhance model resilience against adversarial inputs and discussions.

• Compress models to fulfill stringent size and latency requirements without compromising accuracy.

• Identify and resolve issues that arise during training.

• Collaborate with top AI researchers on initiatives focused on training and refining cutting-edge AI systems.

⛳️ Requirements

• Minimum of 3 years of experience in machine learning research; PhD research is applicable towards this requirement.

• Proficient experience with ML frameworks such as PyTorch, JAX, TensorFlow, or similar.

• A degree from a top-100 university, experience at a FAANG or equivalent AI organization, or a comparable research background demonstrated through publications or significant open-source contributions.

• Specialized knowledge in areas including adversarial robustness, efficient computer vision, generative image modeling, post-training of LLMs and behavioral robustness, or multilingual pre-training.

• Experience in adversarial training of image classifiers, robust accuracy evaluation, and managing the trade-offs between robustness and accuracy.

• Proven experience in training image classifiers from start to finish, model compression, and deployment within strict size or latency constraints.

• Experience in training generative image models from scratch and assessing sample quality using metrics like FID.

• Familiarity with supervised fine-tuning and preference optimization of open-weight language models.

• Experience in shaping conversational behavior over multiple interactions while maintaining general capabilities during behavior-specific fine-tuning.

• Experience in training multilingual or low-resource language models from the ground up and designing tokenizers for various scripts.

• Additional experience in scaling laws, curriculum learning, model evaluation, uncertainty estimation, data augmentation, or synthetic data is advantageous.

• Must not require H-1B or STEM OPT sponsorship.

🏝️ Benefits

• Flexible, project-oriented work.

• Attractive compensation package.

• Fully remote working arrangements.

• Ability to work on your own schedule.

• Weekly payments through Stripe or Wise.

• Opportunity to collaborate with leading AI researchers.

• Project timelines can be adjusted based on needs and performance.

• Reasonable accommodations available upon request.

• Referral bonuses up to $480 for each successful referral.