# World-Time Compute with Verified Code World Models

arXiv: https://arxiv.org/abs/2609.09163 · HTML: https://mlsanity.com/paper/2609.09163/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in World models for language agents, one of 11 directions in World models.
- Builds on 45 of its 53 references that are in the map.
- First posted on arXiv 7 Jul 2026.

## Placed in

- World models / [World models for language agents](https://mlsanity.com/wm/world-models-for-language-agents/)

## Builds on (most cited first, 30 of 45)

- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Optuna: A Next-generation Hyperparameter Optimization Framework](https://mlsanity.com/paper/1907.10902/) (2019)
- [Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models](https://mlsanity.com/paper/2206.04615/) (2022)
- [Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters](https://mlsanity.com/paper/2408.03314/) (2024)
- [World Models](https://mlsanity.com/paper/1803.10122/) (2018)
- [Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World](https://mlsanity.com/paper/1703.06907/) (2017)
- [Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model](https://mlsanity.com/paper/1911.08265/) (2019)
- [Mastering Diverse Domains through World Models](https://mlsanity.com/paper/2301.04104/) (2023)
- [Tulu 3: Pushing Frontiers in Open Language Model Post-Training](https://mlsanity.com/paper/2411.15124/) (2024)
- [STaR: Bootstrapping Reasoning With Reasoning](https://mlsanity.com/paper/2203.14465/) (2022)
- [WorldSimBench: Towards Video Generation Models as World Simulators](https://mlsanity.com/paper/2410.18072/) (2024)
- [V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning](https://mlsanity.com/paper/2506.09985/) (2025)
- [Genie: Generative Interactive Environments](https://mlsanity.com/paper/2402.15391/) (2024)
- [Reinforced Self-Training (ReST) for Language Modeling](https://mlsanity.com/paper/2308.08998/) (2023)
- [Leveraging Procedural Generation to Benchmark Reinforcement Learning](https://mlsanity.com/paper/1912.01588/) (2019)
- [Scaling Relationship on Learning Mathematical Reasoning with Large Language Models](https://mlsanity.com/paper/2308.01825/) (2023)
- [Beyond Human Data: Scaling Self-Training for Problem-Solving with Language Models](https://mlsanity.com/paper/2312.06585/) (2023)
- [Transformers are Sample-Efficient World Models](https://mlsanity.com/paper/2209.00588/) (2022)
- [Diffusion for World Modeling: Visual Details Matter in Atari](https://mlsanity.com/paper/2405.12399/) (2024)
- [TTRL: Test-Time Reinforcement Learning](https://mlsanity.com/paper/2504.16084/) (2025)
- [Thinking Fast and Slow with Deep Learning and Tree Search](https://mlsanity.com/paper/1705.08439/) (2017)
- [Agent-as-a-Judge: Evaluate Agents with Agents](https://mlsanity.com/paper/2410.10934/) (2024)
- [How Far is Video Generation from World Model: A Physical Law Perspective](https://mlsanity.com/paper/2411.02385/) (2024)
- [ARC-AGI-2: A New Challenge for Frontier AI Reasoning Systems](https://mlsanity.com/paper/2505.11831/) (2025)
- [Open-Ended Learning Leads to Generally Capable Agents](https://mlsanity.com/paper/2107.12808/) (2021)
- [Evaluating the World Model Implicit in a Generative Model](https://mlsanity.com/paper/2406.03689/) (2024)
- [WorldCoder, a Model-Based LLM Agent: Building World Models by Writing Code and Interacting with the Environment](https://mlsanity.com/paper/2402.12275/) (2024)
- [The CLRS Algorithmic Reasoning Benchmark](https://mlsanity.com/paper/2205.15659/) (2022)
- [ARC Prize 2024: Technical Report](https://mlsanity.com/paper/2412.04604/) (2024)

## Built on by (most cited first, 1 of 1)

- [Planetary Geospatial Foundation Models: A New Paradigm for Global Public Health](https://mlsanity.com/paper/2610.05699/) (2026)

## Abstract (arXiv)

LLMs generalize across a domain only after seeing many real, labeled examples, which most domains lack. We study a way to manufacture it cheaply. When a domain's dynamics can be written as code, one template instantiates into many world models: executable, verifiable programs over symbolic state, each an inexhaustible source of exactly-labeled trajectories. Fine-tuning an LLM on trajectories through many such worlds, which we call world-time compute, a training-time analogue of test-time compute, lifts generalization to held-out worlds it never trained on (synthesized world families). Gains are largest where capability is scarcest: +29 points at 0.5B; the largest model's lift is within noise, consistent with saturation. Labels can be trusted because the worlds are verified code: synthesized-then-checked dynamics are exact over 20-step rollouts and answer 10x out-of-distribution probes exactly (100%), whereas per-step LLM and MLP predictors compound error and collapse. Unlike domain randomization, each world is independently authored and verified; a corrupted-label control shows label exactness, not task variety, drives the gains. On real benchmarks (ARC-AGI grids, List Functions, CLRS) the same lever holds as per-world test-time training. On List Functions the harder cross-world form holds: one adapter trained on 128 disjoint worlds reaches 40% on held-out worlds versus 6% for a corrupted-label control (+34 points, CI [29, 39]). The gain is a saturating regularity, not a law: largest for few-step reasoning and small/weak models, fading for long chains, perception-induced tasks, and saturated tasks; cross-task transfer is weak without shared skill. Worlds are authored and served by OpenWorld, a zero-dependency framework (companion paper). Scope: symbolic state; pixel-native domains remain territory of learned models. All code, recipes, and this manuscript regenerate from one repository.