Staff Research Scientist - RL Post-Training for Video Gen

Sarah Smith Fund; Palo Alto (CA); On-site; USD 180,000 - 260,000; Full time

Benefits offered by this job

Health benefits

401k matching

Equity

Job summary

Pika is seeking Research Scientists with expertise in RL post-training and generative modeling for large-scale video generation. You will own RL-based post-training for video diffusion/flow-matching models and help build robust reward models, collaborating with engineering and product teams.

The role covers evaluating models with human and automated metrics, distilling RL-tuned models for efficient deployment, and advancing state-of-the-art video generation capabilities in a fast-moving startup

Qualifications

• 2+ years hands-on research in post-training or generative modeling.
• Experience with RL or preference optimization on generative models.
• Strong grounding in diffusion/flow-matching models and PyTorch.

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/configure preference data collection workflows, train and validate learned judges, and 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

Post-training research

Generative modeling

RL

PyTorch

Distributed training

Tools

Diffusion models

Flow-matching

Video generation

Job description

Pika is seeking Research Scientists with expertise in RL post-training and generative modeling for large-scale video generation. You will own RL-based post-training for video diffusion/flow-matching models and help build robust reward models, collaborating with engineering and product teams.

The role covers evaluating models with human and automated metrics, distilling RL-tuned models for efficient deployment, and advancing state-of-the-art video generation capabilities in a fast-moving startup