Emergent Tangled Graph Representations for Atari Game Playing Agents

51 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.

Revisiting the Arcade Learning Environment: Evaluation Protocols and Open Problems for General Agents (2017, methodology): “Other directions the research community has been looking at are the development of better visualization methods (Zahavy, BenZrihem, & Mannor, 2016), the proposal of algorithms that alleviate the need for specialized hardware (Mnih et al., 2016), and genetic algorithms (Kelly & Heywood, 2017).”
Evolving simple programs for playing atari games (2018, methodology): “The ability to generalize over multiple Atari games was further demonstrated in Kelly and Heywood [10], which followed Kelly and Heywood [9].”
Automatic programming: The open issue? (2019, methodology): “…to allow comparison to, and perhaps hybridisation with, a wider set of machine learning algorithms is the Arcade Learning Environment (ALE) https ://githu b.com/mgbel lemar e/Arcad e-Learn ing-Envir onmen t, which was recently tackled using Cartesian GP [68] and Tangled Program Graph GP [30].”
Text-based adventures of the golovin AI agent (2017, methodology): “In some sense, this approach is coherent with the experiments on learning Atari 2600 games using the Arcade Learning Environment (ALE), where the agent’s inputs were only raw screen capture and a score counter [11], [12].”