ByteDance
Research Scientist- Foundation Model, Vision and Language
ByteDance, Seattle, Washington, us, 98127
Research Scientist- Foundation Model, Vision and Language
Responsibilities Conduct cutting-edge research and development in computer vision and natural language processing, with emphasis on multi-modality, vision and language. Enhance multimodal understanding and reasoning (images and videos) across data acquisition, model evaluation, pre-training, SFT, reward modeling, and reinforcement learning to improve overall performance. Synthesize large-scale, high-quality multi-modal data through rewriting, augmentation, and generation to advance foundation models during pretraining, SFT, and RLHF stages. Investigate and implement robust evaluation methodologies to assess model performance across diverse multimodal skills, understand underlying capabilities, and drive improvements. Qualifications Experience in computer vision and natural language processing, including multi-modal understanding, vision-and-language, multimodal pre-training, visual instruction tuning, alignment learning, and related topics. Experience with very large-scale datasets and building datasets to scale foundation models. Experience with language models and applying them to downstream tasks. Strong programming skills in Python and familiarity with popular deep learning frameworks; solid algorithmic foundation. Ability to collaborate effectively with team members and work independently with strong communication skills. Preferred: publications in top-tier venues (e.g., CVPR, ECCV, ICCV, NeurIPS, ICLR, ICML, EMNLP, ACL, NAACL) and impactful open-source projects on GitHub demonstrated engineering ability to solve new challenges. Job Information (For Pay Transparency) Compensation Description (Annually) The base salary range for this position in the selected city is $177,688 – $341,734 annually. Compensation may vary based on qualifications, skills, competencies, experience, and location. Base pay is part of the Total Package and the role may be eligible for additional bonuses, incentives, and restricted stock units. Benefits may vary by employment type and location. Employees have day-one access to medical, dental, and vision insurance; a 401(k) with company match; paid parental leave; disability and life insurance; wellbeing benefits; and paid time off (e.g., 10 holidays, 10 sick days, 17 days of Paid Personal Time, prorated on hire with tenure-based accruals). The company reserves the right to modify or change benefits at any time, with or without notice. About Doubao (Seed) Founded in 2023, the ByteDance Doubao (Seed) Team pioneers advanced AI foundation models, focusing on cutting-edge research and leadership in AI, with labs in China, Singapore, and the US. Why Join ByteDance ByteDance aims to inspire creativity and enrich life through innovative products and diverse teams. We foster curiosity, humility, and impact, operating with an "Always Day 1" mindset to achieve breakthroughs for our company and users. Join us to create and grow together. Diversity & Inclusion ByteDance is committed to an inclusive space where employees are valued for their skills and perspectives. We celebrate diversity and strive to reflect the communities we serve. Reasonable Accommodation ByteDance provides reasonable accommodations in recruitment for candidates with disabilities or other protected reasons. If you need assistance, please reach out to us at https://tinyurl.com/RA-request ByteDance is a global incubator of platforms at the cutting edge of commerce, content, entertainment, and enterprise services with billions of users engaging with our products.
#J-18808-Ljbffr
Responsibilities Conduct cutting-edge research and development in computer vision and natural language processing, with emphasis on multi-modality, vision and language. Enhance multimodal understanding and reasoning (images and videos) across data acquisition, model evaluation, pre-training, SFT, reward modeling, and reinforcement learning to improve overall performance. Synthesize large-scale, high-quality multi-modal data through rewriting, augmentation, and generation to advance foundation models during pretraining, SFT, and RLHF stages. Investigate and implement robust evaluation methodologies to assess model performance across diverse multimodal skills, understand underlying capabilities, and drive improvements. Qualifications Experience in computer vision and natural language processing, including multi-modal understanding, vision-and-language, multimodal pre-training, visual instruction tuning, alignment learning, and related topics. Experience with very large-scale datasets and building datasets to scale foundation models. Experience with language models and applying them to downstream tasks. Strong programming skills in Python and familiarity with popular deep learning frameworks; solid algorithmic foundation. Ability to collaborate effectively with team members and work independently with strong communication skills. Preferred: publications in top-tier venues (e.g., CVPR, ECCV, ICCV, NeurIPS, ICLR, ICML, EMNLP, ACL, NAACL) and impactful open-source projects on GitHub demonstrated engineering ability to solve new challenges. Job Information (For Pay Transparency) Compensation Description (Annually) The base salary range for this position in the selected city is $177,688 – $341,734 annually. Compensation may vary based on qualifications, skills, competencies, experience, and location. Base pay is part of the Total Package and the role may be eligible for additional bonuses, incentives, and restricted stock units. Benefits may vary by employment type and location. Employees have day-one access to medical, dental, and vision insurance; a 401(k) with company match; paid parental leave; disability and life insurance; wellbeing benefits; and paid time off (e.g., 10 holidays, 10 sick days, 17 days of Paid Personal Time, prorated on hire with tenure-based accruals). The company reserves the right to modify or change benefits at any time, with or without notice. About Doubao (Seed) Founded in 2023, the ByteDance Doubao (Seed) Team pioneers advanced AI foundation models, focusing on cutting-edge research and leadership in AI, with labs in China, Singapore, and the US. Why Join ByteDance ByteDance aims to inspire creativity and enrich life through innovative products and diverse teams. We foster curiosity, humility, and impact, operating with an "Always Day 1" mindset to achieve breakthroughs for our company and users. Join us to create and grow together. Diversity & Inclusion ByteDance is committed to an inclusive space where employees are valued for their skills and perspectives. We celebrate diversity and strive to reflect the communities we serve. Reasonable Accommodation ByteDance provides reasonable accommodations in recruitment for candidates with disabilities or other protected reasons. If you need assistance, please reach out to us at https://tinyurl.com/RA-request ByteDance is a global incubator of platforms at the cutting edge of commerce, content, entertainment, and enterprise services with billions of users engaging with our products.
#J-18808-Ljbffr