Generalized Rapid Action Value Estimation in Memory-Constrained Environments
Aloïs Rautureau, Tristan Cazenave, Éric Piette · 2026
Read Generalized Rapid Action Value Estimation in Memory-Constrained Environments by Aloïs Rautureau, Tristan Cazenave, Éric Piette (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-23318.txt.
Description
Aloïs Rautureau and two co-authors extend GRAVE, a strong Monte-Carlo tree search variant for general game playing, for memory-limited settings. GRAVE's strength comes at a cost: it stores extra win and visit statistics at every node, which makes it impractical where memory is tight and limits its real-world use.
Three new algorithms address the cost in different ways. GRAVE2 uses two-level search, GRAVER recycles nodes, and GRAVER2 combines both techniques. The enhancements drastically cut the number of stored nodes while matching GRAVE's playing strength. For practitioners, the result is a practical route to GRAVE-level play on devices and settings where memory, not compute, is the binding constraint.
Catalog record
| CREATOR | Aloïs Rautureau, Tristan Cazenave, Éric Piette |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-23318