Internship Position
2050 Research Singapore welcome you to become a part of our team as a research intern for multi-modal content generation. Skywork AI's 2050 Research Lab is at the forefront of fundamental research in AGI. This is a great opportunity to put theory into practice, learn from top researchers, and contribute to a groundbreaking R&D landscape.
Conduct cutting-edge research in multi-modal content generation (Diffusion model for generating image/audio/video, etc), and content understanding (MLLM for understanding image/video/audio, etc) ensuring findings are published in prestigious conferences and journals.
Identify and tackle complex problems in multi-modal content generation/understanding under the guidance of expert mentors.
Develop algorithms, insights, or data collections that address key aspects of higher research objectives.
Utilize and contribute to cutting-edge tools and methods in multi-modal content generation/understanding, applying these advancements to product development.
Preference for candidates who can commit to a full-time internship of at least 6 months.
Engaged in a PhD program in computer science or a related STEM discipline, with a focus on image/video/audio generation (Diffusion, etc.) or content understanding (LLM, MLLM, etc.)
Demonstrated excellence in research with publications in top-tier conferences or journals (NeurIPS, ICLR, CVPR, ICCV, ECCV, ICML, or equivalent).
Exhibit a solid understanding of generative models, where expertise in image generation is considered a plus.
Exhibit a solid knowledge of (multimodal) large language model, where experience on LLM pretrained is considered a big plus.
Demonstrate proficient coding skills in deep learning frameworks, like PyTorch and TensorFlow.
Interested candidates may send the latest Curriculum Vitae and Cover Letter to michelle@kunlun-inc.com to explore this exciting opportunity. We are committed to fostering diversity and welcomes candidates from all backgrounds to apply. Join us in shaping the future of AGI at Skywork AI's 2050 Research Lab.
