Evolving a diversity of virtual creatures through novelty search and local competition

By Joel Lehman, Kenneth O. Stanley. 2011

522 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Diversity is All You Need: Learning Skills without a Reward Function (2018, background): “Prior work in neuroevolution and evolutionary algorithms has studied how complex behaviors can be learned by directly maximizing diversity [25, 26, 27, 28, 29, 30, 31].”
Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions (2019, background): “effectively co-evolve, novelty search with local competition [33], MAP-Elites [34, 35], Innovation Engines [36], which exploit the opportunity to transfer high-quality solutions from one objective among many to another, and the combinatorial multi-objective evolutionary algorithm (CMOEA), which exte”
Go-Explore: a New Approach for Hard-Exploration Problems (2019, methodology): “The idea of preserving and exploring from stepping stones in an archive comes from the quality diversity (QD) family of algorithms (like MAP-elites [60, 79] and novelty search with local competition [80]), and Go-Explore is an enhanced QD algorithm based on MAP-Elites.”
Enhanced POET: Open-Ended Reinforcement Learning through Unbounded Invention of Learning Challenges and their Solutions (2020, background): “…(e.g. pursuing different kinds of solutions simultaneously) and optimization (e.g. giving one arguably “best” solution) (Brant & Stanley, 2017; Lehman & Stanley, 2011b; Mouret & Clune, 2015; Pugh et al., 2016; Soros & Stanley, 2014), as those solely focusing on optimization often lead to…”
A Recipe for Soft Fluidic Elastomer Robots (2015, unknown): “6 VoxCAD and the Voxelyze physics engine, as used in Cheney et al. 61 and Lehman and Stanley, 62 and reviewed by Lipson, 6 are simulation tools for very soft nonlinear materials.”
Challenges and Opportunities for Design, Simulation, and Fabrication of Soft Robots (2014, unknown): “4,9 In nonlinear relaxation, the structure is represented as a network of simple elements such as springs, beams, and masses.”
AI-GAs: AI-generating algorithms, an alternate paradigm for producing general artificial intelligence (2019, methodology): “The ﬁrst of its kind was Novelty Search with Local Competition [93] followed by MAP-Elites [116].”
Improving Exploration in Evolution Strategies for Deep Reinforcement Learning via a Population of Novelty-Seeking Agents (2017, methodology): “For the Humanoid Locomotion problem the BC is the agent’s ﬁnal { x, y } location, as it was in Lehman & Stanley (2011c).”
Self-Supervised Exploration via Disagreement (2019, methodology): “…al., 2006) of less-frequently visited states, “curiosity” (Oudeyer & Kaplan, 2009; Pathak et al., 2017; Schmidhuber, 1991a) where prediction error is used as reward signal and “diversity rewards” (Eysenbach et al., 2018; Lehman & Stanley, 2011a;b) which incentivize diversity in the visited states.”
Soft Manipulators and Grippers: A Review (2016, methodology): “Some simulators have been developed using non-linear relaxation for kinematic simulation, whereby the system is represented by a system of springs, beams, and masses (Lehman and Stanley, 2011; Cheney et al., 2013).”
Learning Novel Policies For Tasks (2019, background): “In the same spirit of encouraging diversity over an entire trajectory, work from the evolutionary algorithms community has applied Novelty Search (Lehman & Stanley, 2011a;b; Pugh et al., 2016; Mouret & Doncieux, 2009) to solve various problems.”
Minimal criterion coevolution: a new approach to open-ended search (2017, background): “QD algorithms such as novelty search with local competition (NSLC) [11] and the Multi-dimensional Archive of Phenotypic Elites (MAP-Elites) [17] have produced arguably practical results, such as a repertoire of robot walking strategies [29], even as they incorporate some of the flavor of alife…”
Illuminating search spaces by mapping elites (2015, methodology): “In 2011, Lehman and Stanley 22 note that combining a selective pressure for feature diversity with one performance objective that all of the different types of phenotypes compete on is akin to having butterﬂies and bacteria compete with bears and badgers on one performance criterion (e.g. speed).”
Planning to Explore via Self-Supervised World Models (2020, background): “…that it cannot yet predict accurately (Schmid-huber, 1991b; Oudeyer et al., 2007; Pathak et al., 2017), maximally inﬂuence its inputs (Klyubin et al., 2005; Eysen-bach et al., 2018), or visit rare states (Poupart et al., 2006; Lehman & Stanley, 2011; Bellemare et al., 2016; Burda et al., 2018).”