Google DeepMind is looking to enhance Gemini's capabilities in code security analysis by designing and implementing advanced post-training strategies.
Requirements
- Experiences in fine-tuning and adaptation of LLMs (e.g. advanced prompting, supervised fine-tuning, RLHF)
- Strong knowledge of systems design and data structures
- Proven experience with TensorFlow, JAX, PyTorch, or similar leading deep learning frameworks
- Recent experience conducting applied research to improve the quality and training/serving efficiency of large transformer-based models
- Deep understanding of statistics is strongly preferred
Responsibilities
- Design and Implement advanced post-training algorithms (SFT, RLHF, RLAIF) to optimize Gemini for code security tasks and secure coding practices.
- Diagnose and interpret training outcomes (regressions in coding ability, gains in security reasoning), and propose solutions to improve model capabilities.
- Actively monitor and evolve the system's performance through metric design.
- Develop reliable automated evaluation pipelines for code security that are strongly correlated with human security expert judgment.
- Construct complex benchmarks to probe the limits of the model’s ability to reason about control flow, memory safety, and software weakness.
Other
- BSc, MSc or PhD/DPhil degree in computer science, stats, machine learning or similar experience working in industry
- A passion for Artificial Intelligence.
- Excellent communication skills and proven interpersonal skills, with a track record of effective collaboration with cross-functional teams
- We are seeking individuals who excel in fast-pacing environments and are eager to contribute to the advancement of AI.
- We highly value the ability to invent novel solutions to complex problems, embracing a can-do and fail-fast mindset.
