Sotabase

Career

· Member Of Technical Staff, Microsoft AI2025–
· Research Engineer, DeepMind2017–2024
· Research Associate, University of Cambridge2016–2017
· Doctor of Philosophy (PhD) Computer Science, University of Cambridge2012–2016
· Doctor of Philosophy (PhD), Computer Science, University of Cambridge2012–2016
· PHD Student, University of Cambridge2012–2016
· Master’s Degree, Computer Science, Distinction, University of Cambridge2008–2012

Publications (22)

Grandmaster level in StarCraft II using multi-agent reinforcement learning
Nature · 2019
4,163
cited
Population Based Training of Neural Networks
arXiv.org · 2017
828
cited
Open-Ended Learning Leads to Generally Capable Agents
arXiv.org · 2021
220
cited
RoboCat: A Self-Improving Foundation Agent for Robotic Manipulation
arXiv.org · 2023
110
cited
BOAT: Building Auto-Tuners with Structured Bayesian Optimization
The Web Conference · 2017
88
cited
RoboCat: A Self-Improving Generalist Agent for Robotic Manipulation
2023
80
cited
A Generalized Framework for Population Based Training
Knowledge Discovery and Data Mining · 2019
70
cited
Rapid training of deep neural networks without skip connections or normalization layers using Deep Kernel Shaping
arXiv.org · 2021
52
cited
Discovering Attention-Based Genetic Algorithms via Meta-Black-Box Optimization
Annual Conference on Genetic and Evolutionary Computation · 2023
40
cited
PrefEdge: SSD Prefetcher for Large-Scale Graph Traversal
Annual Haifa Experimental Systems Conference · 2014
31
cited
Faster Improvement Rate Population Based Training
arXiv.org · 2021
14
cited
A framework to build bespoke auto-tuners with structured Bayesian optimisation
2017
12
cited
Learning Runtime Parameters in Computer Systems with Delayed Experience Injection
arXiv.org · 2016
10
cited
Community Detection in Multi-Layer Networks
2012
4
cited
Mitigating I/O latency in SSD-based graph traversal
2012
2
cited
Tuning the Scheduling of Distributed Stochastic Gradient Descent with Bayesian Optimization
arXiv.org · 2016
2
cited
Perception-Prediction-Reaction Agents for Deep Reinforcement Learning
arXiv.org · 2020
1
cited
Population Based Training as a Service
2018
StructMoE : Augmenting MoEs with Hierarchically Routed Low Rank Experts
Towards Policy-Guided Conversational Recommendation with Dialogue Acts
2022
Sotabase
Valentin Dalibard | Researcher Profile | Sotabase | Sotabase