![[2602.23318] Generalized Rapid Action Value Estimation in Memory-Constrained Environments](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.23318] Generalized Rapid Action Value Estimation in Memory-Constrained Environments
Summary
The paper presents GRAVE2, GRAVER, and GRAVER2, enhanced algorithms for Generalized Rapid Action Value Estimation, addressing memory constraints in Monte-Carlo Tree Search applications.
Why It Matters
This research is significant as it improves the efficiency of game-playing algorithms in memory-constrained environments, expanding their practical applicability. By optimizing memory usage while maintaining performance, it opens new avenues for AI applications in resource-limited settings.
Key Takeaways
- Introduction of GRAVE2, GRAVER, and GRAVER2 algorithms.
- Enhancements focus on two-level search and node recycling.
- Significant reduction in stored nodes while preserving performance.
- Addresses practical limitations of existing algorithms in memory-constrained environments.
- Contributes to advancements in General Game Playing (GGP) AI.
Computer Science > Artificial Intelligence arXiv:2602.23318 (cs) [Submitted on 26 Feb 2026] Title:Generalized Rapid Action Value Estimation in Memory-Constrained Environments Authors:Aloïs Rautureau, Tristan Cazenave, Éric Piette View a PDF of the paper titled Generalized Rapid Action Value Estimation in Memory-Constrained Environments, by Alo\"is Rautureau and 2 other authors View PDF HTML (experimental) Abstract:Generalized Rapid Action Value Estimation (GRAVE) has been shown to be a strong variant within the Monte-Carlo Tree Search (MCTS) family of algorithms for General Game Playing (GGP). However, its reliance on storing additional win/visit statistics at each node makes its use impractical in memory-constrained environments, thereby limiting its applicability in practice. In this paper, we introduce the GRAVE2, GRAVER and GRAVER2 algorithms, which extend GRAVE through two-level search, node recycling, and a combination of both techniques, respectively. We show that these enhancements enable a drastic reduction in the number of stored nodes while matching the playing strength of GRAVE. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2602.23318 [cs.AI] (or arXiv:2602.23318v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2602.23318 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Aloïs Rautureau [view email] [v1] Thu, 26 Feb 2026 18:25:59 UTC (338 KB) Full-text links: Access Paper: View a PDF of th...