Jumanji scalable compositional JAX environments

What changed

Provides accelerator-native routing/packing/logic environments with customizable initial-state generators, reset/reward/viewer components and actor-critic baselines for scalable experiments.

Scientific scope

Provides accelerator-native routing/packing/logic environments with customizable initial-state generators, reset/reward/viewer components and actor-critic baselines for scalable experiments. Original v1 has22authors and18environments, not the later official site's24authors/22environments. NP-hard problem environments do not imply polynomial-time exact solvers. Hardware acceleration and selected generalization experiments do not establish arbitrary real-world transfer. First-author star is correspondence.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Original v1 has22authors and18environments, not the later official site's24authors/22environments. NP-hard problem environments do not imply polynomial-time exact solvers. Hardware acceleration and selected generalization experiments do not establish arbitrary real-world transfer. First-author star is correspondence.

Contributors

• Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.

• Jumanji: a Diverse Suite of Scalable Reinforcement Learning Environments in JAX

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction