# Chess Simulation ## Summary This application provides a Chess environment for RL agents, supporting both numeric and language-based state representations. Agents can learn to play chess using various adapters and configurations, with support for LLM-based state descriptions. ## Data Source The Chess environment and adapters are based on open-source implementations and extended for use with elsciRL. ## elsciRL App Options ### Local Config | Config Name | Config Code | Description | |:-----------:|:-----------:|:-----------| | Default | default | Standard chess configuration with all pieces and standard rules. | ### Adapters | Adapter Name | Adapter Code | Description | Output type | |:-----------------:|:------------:|:---------------------------------------------------------------------------:|:-----------:| | Numeric Board | numeric | Encodes the board as a tensor of piece indices. | $tensor$ | | Language Active | language | Describes the board in natural language, focusing on active pieces. | $tensor$ | | LLM Adapter | llm | Uses an LLM to generate a language description of the board state. | $tensor$ | ## Citation ```bibtex @phdthesis{Osborne2024, title = {Improving Real-World Reinforcement Learning by Self Completing Human Instructions on Rule Defined Language}, author = {Philip Osborne}, year = 2024, month = {August}, address = {Manchester, UK}, note = {Available at \url{https://research.manchester.ac.uk/en/studentTheses/improving-real-world-reinforcement-learning-by-self-completing-hu}}, school = {The University of Manchester}, type = {PhD thesis} } ```