Senior RL Video Gen Scientist — Hybrid, Equity

Sarah Smith Fund; United States; Hybrid; USD 150,000 - 230,000; Full time

Benefits offered by this job

Health benefits

Equity

401k matching

Growth opportunities

Hybrid on-site/remote

Job summary

Pika is seeking Research Scientists to advance RL alignment and generative modeling for real-time video generation. You will own RL post-training for video diffusion/flow-matching models, develop reward models, and guide evaluation across human and automated metrics.

You will collaborate with engineering and product teams, work on multi-node distributed training, and help distill RL-tuned models into efficient few-step samplers while preserving alignment gains. This role is staff/lead level.

Qualifications

• 2+ years hands-on research in post-training or generative modeling.
• RL or preference-optimization experience on generative models with evidence of model improvement.
• Strong grounding in diffusion or flow-matching models, PyTorch, and multi-node distributed training.

Responsibilities

• Run RL post-training (preference optimization, online RL against learned rewards) for video diffusion/flow-matching models at multi-node scale.
• Build video reward models: define target evaluation dimensions, design data collection workflows, train and validate learned judges, safeguard against reward hacking.
• Own post-training evaluation, including human preference studies and their correlation with automated metrics.
• Distill RL-tuned models to efficient few-step samplers while preserving alignment gains (secondary focus).

Skills

RL post-training

Generative modeling

PyTorch

Multi-node distributed training

Tools

PyTorch

Job description

Pika is seeking Research Scientists to advance RL alignment and generative modeling for real-time video generation. You will own RL post-training for video diffusion/flow-matching models, develop reward models, and guide evaluation across human and automated metrics.

You will collaborate with engineering and product teams, work on multi-node distributed training, and help distill RL-tuned models into efficient few-step samplers while preserving alignment gains. This role is staff/lead level.