Staff Scientist: RL & Reward Modeling for Diffusion LLMs

Inception; San Francisco (CA); On-site; USD 180,000 - 260,000; Full time

Job summary

Inception seeks experienced scientists and engineers with deep expertise in post-training large language models through reinforcement learning. You will design and implement RL training pipelines for diffusion LLMs, develop reward modeling strategies, and build the algorithms that align model behavior with human intent at scale.

The role focuses on designing RL training pipelines, building reward models, and advancing alignment techniques for enterprise-grade LLMs, with emphasis on stability,

Qualifications

• PhD or equivalent research/engineering experience in ML.
• Proficient with PyTorch and distributed training.
• Strong knowledge of transformers, LLMs, and RLHF workflows.

Responsibilities

• Design, develop, and optimize RL training pipelines (PPO, DPO, RLHF) for diffusion-based LLMs.
• Build and iterate reward models and evaluation of reward quality.
• Implement approaches for fine-tuning and scaling generative AI models.
• Work on data preprocessing, model evaluation, and enterprise alignment.
• Research techniques for controlled text generation and constraint satisfaction.
• Improve training stability, efficiency, and reproducibility of RL workloads.

Skills

RLHF experience

PPO/DPO familiarity

Transformer models

Distributed training

PyTorch experience

Education

BS/MS/PhD in CS

Tools

vLLM

TensorRT

SGLang

Job description

Inception seeks experienced scientists and engineers with deep expertise in post-training large language models through reinforcement learning. You will design and implement RL training pipelines for diffusion LLMs, develop reward modeling strategies, and build the algorithms that align model behavior with human intent at scale.

The role focuses on designing RL training pipelines, building reward models, and advancing alignment techniques for enterprise-grade LLMs, with emphasis on stability,