Lead RL Post-Training Scientist: Video Generation

pika; United States; Hybrid; USD 180,000 - 280,000; Full time

Benefits offered by this job

Competitive salary

Full health benefits

401k matching

Hybrid work model (HQ in Palo Alto)

Job summary

Pika is seeking Research Scientists to lead RL‑driven post‑training for real‑time video generation models and to develop reward models that align with human preferences. You will collaborate across engineering and product to push the frontier of agentic, multimodal video platforms.

This staff/lead role focuses on refining diffusion/flow‑matching models, distilling RL‑tuned models, and evaluating results with both human studies and automated metrics.

Qualifications

• 2+ years hands‑on research experience 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

Hands-on RL post-training

RL / preference-optimization

Diffusion / flow-matching

Tools

PyTorch

Job description

Pika is seeking Research Scientists to lead RL‑driven post‑training for real‑time video generation models and to develop reward models that align with human preferences. You will collaborate across engineering and product to push the frontier of agentic, multimodal video platforms.

This staff/lead role focuses on refining diffusion/flow‑matching models, distilling RL‑tuned models, and evaluating results with both human studies and automated metrics.