Nicolas Sonnerat
Google DeepMind (United Kingdom) · 19 works · 2,918 citations · h-index 11
Top works
- Gemini: A Family of Highly Capable Multimodal Models
- Human-level performance in 3D multiplayer games with population-based reinforcement learning
- Value-Decomposition Networks For Cooperative Multi-Agent Learning Based On Team Reward
- Learning a Generic Value-Selection Heuristic Inside a Constraint Programming Solver
- Deep Reinforcement Learning and the Deadly Triad
- SCAN: Learning Hierarchical Compositional Visual Concepts
- Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2
- Solving Mixed Integer Programs Using Neural Networks
- SCAN: Learning Abstract Hierarchical Compositional Visual Concepts
- Olympiad-level formal mathematical reasoning with reinforcement learning
- Galaxy cutsets in graphs
- Learning a Large Neighborhood Search Algorithm for Mixed Integer Programs
- Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders
- Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2
- Maximum Flows on Disjoint Paths
- Defending Planar Graphs against Star-Cutsets