Parameter Optimization & Emergent Behavior in DeFi by Tom McLean, March Sabate-Vidales & David Siska

Emergent Behavior in DeFi: Agent Based Simulations and Reinforcement Learning As in any complex system there may be emergent behaviors in DeFi protocols. In this workshop we will show the basics of how agent based simulations combined with reinforcement learning can be used to explore these and also for optimization of various protocol parameter values. We will begin with an introduction to Vega Protocol itself before moving on to discuss how we have used agent-based modelling to both test the protocol itself and investigate behaviour under differing conditions to inform design and configuration decisions. From this baseline, we will then guide participants in setting up and running preconfigured simulations with Vega Protocol. Once we are all set up, and depending upon audience interests, we can then work through more custom setups with either a focus on Reinforcement Learning or testing trading strategies in a realistic living market. This workshop will be based on https://github.com/vegaprotocol/vega-market-sim with additional references to https://github.com/msabvid/cpm_agent_based_sim. Working through the setup on the Vega Market Sim repo will give a great headstart in the workshop. Tom McLean March Sabate-Vidales David Siska