Video Foundation Model Research Lead

LIGHTSPEED STUDIOS; Singapore; On-site; SGD 120,000 - 180,000; Full time

Job summary

LIGHTSPEED STUDIOS is seeking a PhD-level researcher to advance video generation foundations. You will design architecture, study long-context attention, and optimize memory mechanisms for high-quality, scalable video synthesis.

Responsibilities include leading pre-training at large scale, exploring tokenizers, and driving scientific roadmap decisions while publishing results in top conferences. Experience with distributed training and multi-resolution models is preferred.

Qualifications

• Ph.D. in AI-related fields with first-author papers at top-tier conferences.
• Proficient in the principles and engineering implementation of diffusion models and autoregressive generation.
• Experience training video/image generation models from scratch; highly proficient in PyTorch and large-scale distributed training.
• Deep understanding of the design trade-offs in Video VAE/Tokenizers, with 3+ years of relevant research experience.
• Publications related to video generation or diffusion models.
• Experience leading end-to-end video foundation model projects is preferred.

Responsibilities

• Design and iterate on the next-generation video generation foundation architecture.
• Explore core technologies for long video generation, including long-context attention, KV Cache compression, and memory mechanisms.
• Research high-compression-ratio video tokenizers and advance unified modeling capabilities for multi-resolution and multi-frame-rate videos.
• Lead video pre-training at large token scale, defining data mixture and curriculum learning strategies.
• Explore Scaling Laws, define scientific scale-up paths, and continuously improve key model capabilities.
• Track industry state-of-the-art, lead comparative experiments, drive technical roadmap decisions, and publish academic papers.

Skills

PyTorch

Diffusion models

Autoregressive generation

Distributed training

Video generation

Education

Ph.D. in AI-related field

Tools

High-performance GPUs

Job description

LIGHTSPEED STUDIOS is seeking a PhD-level researcher to advance video generation foundations. You will design architecture, study long-context attention, and optimize memory mechanisms for high-quality, scalable video synthesis.

Responsibilities include leading pre-training at large scale, exploring tokenizers, and driving scientific roadmap decisions while publishing results in top conferences. Experience with distributed training and multi-resolution models is preferred.