ARCLE: The Abstraction and Reasoning Corpus Learning Environment for Reinforcement Learning

Hosung Lee, Sejin Kim, Seungpil Lee, Sanha Hwang, Jihwan Lee, Byung-Jun Lee, Sundong Kim
Proceedings of The 3rd Conference on Lifelong Learning Agents, PMLR 274:710-731, 2025.

Abstract

This paper introduces ARCLE, an environment designed to facilitate reinforcement learning research on the Abstraction and Reasoning Corpus (ARC). Addressing this inductive reasoning benchmark with reinforcement learning presents these challenges: a vast action space, a hard-to-reach goal, and a variety of tasks. We demonstrate that an agent with proximal policy optimization can learn individual tasks through ARCLE. The adoption of non-factorial policies and auxiliary losses led to performance enhancements, effectively mitigating issues associated with action spaces and goal attainment. Based on these insights, we propose several research directions and motivations for using ARCLE, including MAML, GFlowNets, and World Models.

Cite this Paper

BibTeX
@InProceedings{pmlr-v274-lee25a, title = {ARCLE: The Abstraction and Reasoning Corpus Learning Environment for Reinforcement Learning}, author = {Lee, Hosung and Kim, Sejin and Lee, Seungpil and Hwang, Sanha and Lee, Jihwan and Lee, Byung-Jun and Kim, Sundong}, booktitle = {Proceedings of The 3rd Conference on Lifelong Learning Agents}, pages = {710--731}, year = {2025}, editor = {Lomonaco, Vincenzo and Melacci, Stefano and Tuytelaars, Tinne and Chandar, Sarath and Pascanu, Razvan}, volume = {274}, series = {Proceedings of Machine Learning Research}, month = {29 Jul--01 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v274/main/assets/lee25a/lee25a.pdf}, url = {https://proceedings.mlr.press/v274/lee25a.html}, abstract = {This paper introduces ARCLE, an environment designed to facilitate reinforcement learning research on the Abstraction and Reasoning Corpus (ARC). Addressing this inductive reasoning benchmark with reinforcement learning presents these challenges: a vast action space, a hard-to-reach goal, and a variety of tasks. We demonstrate that an agent with proximal policy optimization can learn individual tasks through ARCLE. The adoption of non-factorial policies and auxiliary losses led to performance enhancements, effectively mitigating issues associated with action spaces and goal attainment. Based on these insights, we propose several research directions and motivations for using ARCLE, including MAML, GFlowNets, and World Models.} }
Endnote
%0 Conference Paper %T ARCLE: The Abstraction and Reasoning Corpus Learning Environment for Reinforcement Learning %A Hosung Lee %A Sejin Kim %A Seungpil Lee %A Sanha Hwang %A Jihwan Lee %A Byung-Jun Lee %A Sundong Kim %B Proceedings of The 3rd Conference on Lifelong Learning Agents %C Proceedings of Machine Learning Research %D 2025 %E Vincenzo Lomonaco %E Stefano Melacci %E Tinne Tuytelaars %E Sarath Chandar %E Razvan Pascanu %F pmlr-v274-lee25a %I PMLR %P 710--731 %U https://proceedings.mlr.press/v274/lee25a.html %V 274 %X This paper introduces ARCLE, an environment designed to facilitate reinforcement learning research on the Abstraction and Reasoning Corpus (ARC). Addressing this inductive reasoning benchmark with reinforcement learning presents these challenges: a vast action space, a hard-to-reach goal, and a variety of tasks. We demonstrate that an agent with proximal policy optimization can learn individual tasks through ARCLE. The adoption of non-factorial policies and auxiliary losses led to performance enhancements, effectively mitigating issues associated with action spaces and goal attainment. Based on these insights, we propose several research directions and motivations for using ARCLE, including MAML, GFlowNets, and World Models.
APA
Lee, H., Kim, S., Lee, S., Hwang, S., Lee, J., Lee, B. & Kim, S.. (2025). ARCLE: The Abstraction and Reasoning Corpus Learning Environment for Reinforcement Learning. Proceedings of The 3rd Conference on Lifelong Learning Agents, in Proceedings of Machine Learning Research 274:710-731 Available from https://proceedings.mlr.press/v274/lee25a.html.

Related Material