LLM Research Scientist

This role focuses on advancing the field of language models through empirical research and hands-on training. As an LLM Research Scientist, you will tackle open-ended research problems, enhancing both pre-training and post-training methodologies for transformer-based language models. Key Responsibilities

• Train transformer-based language models from scratch and fine-tune open-weight models.
• Maximize the efficiency of limited data and compute resources.
• Construct training corpora from raw web-scale sources.
• Build and optimize post-training pipelines.
• Diagnose and resolve training-related issues.

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

• Experience with training transformer-based language models from scratch, end-to-end.
• Ability to operate in data- and compute-constrained environments, effectively allocating fixed budgets across model size, tokens, and epochs.
• Skills in diagnosing optimization failures, convergence issues, and training instabilities.

• Experience in corpus construction from raw web crawls and large unfiltered sources.
• Proficiency in data filtering, deduplication, quality classification, and optimization of mixture/ordering.
• Ability to measure data interventions rigorously.

• Hands-on experience with supervised fine-tuning, including dataset creation through synthetic generation, noisy supervision, and rejection sampling.
• Knowledge of preference optimization (DPO, RLHF, RLAIF) and reward modeling/human-preference prediction.
• Experience in alignment fine-tuning to shape refusal behavior, truthfulness, and unbiased reasoning while maintaining general capabilities.
• Fine-tuning expertise for narrow, verifiable domains (math, code, games, structured prediction) where outputs can be programmatically verified.

• Experience with scaling laws and training-efficiency research.
• Knowledge of curriculum learning and data ordering.
• Experience in LLM evaluation, including benchmark construction and contamination control.
• Familiarity with reinforcement learning for language models.
• Understanding of model alignment and AI safety.

• 3+ years of machine learning research experience (PhD research counts toward this requirement).
• Strong experience with PyTorch, JAX, TensorFlow, or similar ML frameworks.
• A 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.

Employment type is hourly, with flexible, project-based work arrangements.

Hourly compensation ranges from $100 to $120.

This position is remote and open to qualified candidates globally.

• Engage in cutting-edge foundation model research.
• Collaborate with leading AI researchers on high-impact projects.
• Enjoy flexible work arrangements with competitive compensation.