SpellbrushChiyoda City, JapanConduct AI research on large-scale diffusion models for anime imagery and other modalities, moving experimental models rapidly into production. The role requires research experience, diffusion-model knowledge, and PyTorch proficiency, with on-site work in Tokyo or San Francisco.
Build and deploy production LLM agents and internal AI tools, partnering with business teams to turn operational needs into secure, ROI-driven solutions. The role requires 5+ years of software experience, production AI expertise, full-stack capability, and strong cross-functional communication.
Conduct research and build evaluation systems for autonomous AI agents operating on real engineering workflows. The role combines production experimentation, statistical measurement, automated optimization, and post-training open-weight vision-language models using proprietary autonomous-driving data.
Researcher developing evaluations, red-teaming pipelines, and novel mitigations for frontier AI safety risks. The role requires deep technical expertise, research engineering experience, advanced training in computer science or machine learning, and proficiency in Python or similar languages.
Conduct strategic, technically rigorous research to anticipate and mitigate loss-of-control risks from increasingly capable AI systems, including recursive self-improvement. The role combines hypothesis-driven research, rapid prototyping, safety evaluations, monitoring, and institutionalizing effective interventions.
Owns mid-training for LLMs, optimizing data mixes, synthetic data pipelines, annealing schedules, and context extension to enhance reasoning, coding, and math capabilities for AI agents. Requires deep LLM pipeline expertise, hands-on large model training, and original research contributions.