1.Research the model's ability to understand generated content, parsing semantics, objects, relationships, and spatial structures.
2.Implement state tracking and consistency modeling for generated videos.
3.Explore "unified generation-understanding" model architectures.
4.Research causal consistency control and physical constraints for the "action input → video output" pipeline.
5.Research hybrid architectures to achieve low-latency feedback for real-time interactive generation.
6.Collaborate with Agent/RL teams to drive an end-to-end closed loop of "generation → understanding → control → feedback."
7.Validate controllable interactive capabilities in game scenarios and lead the establishment of evaluation systems.
8.Continuously track industry-related work and formulate proprietary technical roadmaps.
1.Ph.D. in AI-related fields, covering areas such as video understanding, video prediction, reinforcement learning, or multimodal generation and understanding.
2.Deep understanding of the internal mechanisms of both video generation and VLM models; familiar with diffusion model principles.
3.Research experience in interactive video generation or controllable video generation.
4.Ability to independently design complex model training pipelines with strong engineering skills.
5.Proficient in Python/PyTorch, with publications in video understanding, controllable generation, or reinforcement learning.
6.Experience with Game AI, simulation environments, or closed-loop systems is preferred.
7.Relevant publications in top-tier conferences or journals are preferred.
As an equal opportunity employer, we firmly believe that diverse voices fuel our innovation and allow us to better serve our users and the community. We foster an environment where every employee of Tencent feels supported and inspired to achieve individual and common goals.