Research Engineer, Gemini Cyber Post-Training, DeepMind

At Google, research-focused Software Engineers are embedded throughout the company, allowing them to setup large-scale tests and deploy promising ideas quickly and broadly. Ideas may come from internal projects as well as from collaborations with research programs at partner universities and technical institutes all over the world.

From creating experiments and prototyping implementations to designing new architectures, engineers work on real-world problems including artificial intelligence, data mining, natural language processing, hardware and software performance analysis, improving compilers for mobile platforms, as well as core search and much more. But you stay connected to your research roots as an active contributor to the wider research community by partnering with universities and publishing papers.

As a Research Engineer in the Cyber Gemini team, you will play a key role in both advancing our research capabilities and building the scalable infrastructure needed to support them. You will work on both the underlying cybersecurity models and custom orchestration harnesses, such as the skills and custom agent loops used for vulnerability verification. You will develop novel datasets and post-training recipes to enable Gemini to become more capable at vulnerability discovery, as well as evaluation datasets to measure the capabilities of our models.

As part of this role, you will also work closely with DeepMind's new Cyber Strike initiative, a high-profile, cross-organizational effort to build cybersecurity models to secure software and enterprise systems at scale. You will actively participate in the finding, fixing, and exploit tracks, curating datasets, building agentic evaluation benchmarks, and integrating CodeMender's vulnerability discovery and patching capabilities into the core Cyber Strike model.

Artificial intelligence will be one of humanity’s most transformative inventions. At DeepMind, we are a pioneering AI lab with exceptional interdisciplinary teams focused on advancing AI development to solve complex global challenges and accelerate high-quality product innovation for billions of users. We use our technologies for widespread public benefit and scientific discovery, ensuring safety and ethics are always our highest priority.

Individual pay is determined by factors including job-related skills, experience, and relevant education or training.

US: $147000 - $210000 (USD) + 15% bonus target + equity + benefits

Learn more about benefits at Google.

Responsibilities

• Build datasets (RL environments, SFT data) and iterate on post-training recipes. 
• Debug and analyze RL training runs to identify potential bottlenecks and optimize training efficiency and effectiveness.
• Build evaluation pipelines, benchmarks, and leaderboards to measure agent performance.

Minimum qualifications:

• PhD degree in Computer Science, or a related field, or equivalent practical experience.
• 2 years of experience (industry or academia) working in Reinforcement Learning environments.
• Experience building distributed system infrastructure.
• One or more publication records in vulnerability research (e.g. auditing, reverse engineering, exploitation).

Preferred qualifications:

• Experience with program analysis techniques (e.g., static analysis, symbolic execution, fuzzing).
• Background in machine learning and statistics.
• Independent, proactive attitude and a passion for the mission of AI safety and security.

Note: By applying to this position you will have an opportunity to share your preferred working location from the following: Mountain View, CA, USA; New York, NY, USA.