LGAIOct 30, 2025

Jasmine: A Simple, Performant and Scalable JAX-based World Modeling Codebase

arXiv:2510.27002v11 citationsh-index: 1
Originality Synthesis-oriented
AI Analysis

This provides scalable and reproducible infrastructure for researchers working on world models, particularly in data-scarce domains like robotics, though it is incremental as it builds on existing methods.

The authors tackled the lack of open training infrastructure for world models by introducing Jasmine, a JAX-based codebase that scales efficiently and achieves an order-of-magnitude faster reproduction of the CoinRun case study compared to prior implementations.

While world models are increasingly positioned as a pathway to overcoming data scarcity in domains such as robotics, open training infrastructure for world modeling remains nascent. We introduce Jasmine, a performant JAX-based world modeling codebase that scales from single hosts to hundreds of accelerators with minimal code changes. Jasmine achieves an order-of-magnitude faster reproduction of the CoinRun case study compared to prior open implementations, enabled by performance optimizations across data loading, training and checkpointing. The codebase guarantees fully reproducible training and supports diverse sharding configurations. By pairing Jasmine with curated large-scale datasets, we establish infrastructure for rigorous benchmarking pipelines across model families and architectural ablations.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes