Senior Interactive Vision Model Researcher

1. Proficient in the principles and engineering implementation of diffusion models and autoregressive generation.

2. Experience training video/image generation models from scratch; highly proficient in PyTorch and large-scale distributed training.

3. Deep understanding of the design trade-offs in Video VAE/Tokenizers, with 3+ years of relevant research experience.

4. Publications related to video generation or diffusion models.

5. Experience in core industry product R&D or hands-on experience with the latest technologies is preferred.

6. Experience leading end-to-end video foundation model projects is preferred.

1. Explore core technologies for long video generation, including long-context attention, KV Cache compression, and memory mechanisms.

2. Research high-compression-ratio video tokenizers and advance unified modeling capabilities for multi-resolution and multi-frame-rate videos.

3. Lead video pre-training at large token scale, defining data mixture and curriculum learning strategies.

4. Explore Scaling Laws, define scientific scale-up paths, and continuously improve key model capabilities.

5. Track industry state-of-the-art , lead comparative experiments, drive technical roadmap decisions, and publish academic papers.

Lightspeed Tech Center is a R&D department under Lightspeed Studios which develop PUBG Mobile and other high-quality games. Our Tech Center leads the research, exploration, and discovery of innovative technologies and provides technical services for all games during all phases of life cycle, including engine, audio, QA, AI, next generation game, technical cooperation, etc.