Research Engineer, Tokens (Pre-training)
Anthropic280,000 - 425,000 USD/year
- Full Time
- Junior (1 to 2 years)
$225,000 – $400,000Compensation
Mid-level (3 to 4 years), Senior (5 to 8 years)Experience Level
Full TimeJob Type
UnknownVisa
AI & Machine Learning, Consumer SoftwareIndustries
Requirements
Candidates must have at least a PhD (or equivalent) and experience writing clear production-facing and training code. A strong understanding of modern machine learning techniques, including reinforcement learning and transformers, is required, along with experience working with GPUs for training, serving, and debugging. Applicants should also have experience with data pipelines and data infrastructure, as well as a track record of exceptional research or creative applied ML projects.
Responsibilities
As a Research Engineer on the Post-Training team, you will develop alignment algorithms and loss functions to enhance data sample efficiency. You will write data pipelines to process diverse web data into a format suitable for models, identify quality signals to evaluate model performance in real-world scenarios, and design sampling algorithms to improve the serving efficiency of large generative models.
Skills
Reinforcement Learning
Transformers
Data Pipelines
GPU Training
Data Infrastructure
Machine Learning
A/B Testing
Kubernetes
Docker
Character.AI
AI platform for creating virtual characters
About Character.AI
Character AI provides a platform that allows users to create and manage virtual characters with distinct personalities. These characters can engage with users in a realistic way, enhancing interactions in various digital settings. The platform is utilized by game developers to create immersive characters that improve gameplay, and by businesses to develop virtual customer service agents that offer a more personalized experience compared to traditional automated systems. Character AI differentiates itself from competitors by focusing on the depth of personality and engagement of its virtual characters. The company aims to capitalize on the growing AI market, providing continuous access to its platform through a subscription or licensing model, which supports its goal of expanding its client base and enhancing digital interactions.
Key Metrics
Menlo Park, CaliforniaHeadquarters
2021Year Founded
$145.9MTotal Funding
SERIES_ACompany Stage
Consumer Software, AI & Machine Learning, GamingIndustries
51-200Employees