# Out-of-Distribution Generalization in the ARC-AGI Domain: Comparing Execution-Guided Neural Program Synthesis and Test-Time Fine-Tuning

arXiv: https://arxiv.org/abs/2507.15877 · HTML: https://mlsanity.com/paper/2507.15877/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 6 of its 33 references that are in the map.
- First posted on arXiv 17 Jul 2025.

## Builds on (most cited first, 6 of 6)

- [Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?](https://mlsanity.com/paper/2504.13837/) (2025)
- [Write, Execute, Assess: Program Synthesis with a REPL](https://mlsanity.com/paper/1906.04604/) (2019)
- [ExeDec: Execution Decomposition for Compositional Generalization in Neural Program Synthesis](https://mlsanity.com/paper/2307.13883/) (2023)
- [Towards Efficient Neurally-Guided Program Induction for ARC-AGI](https://mlsanity.com/paper/2411.17708/) (2024)
- [Execution Guided Line-by-Line Code Generation](https://mlsanity.com/paper/2506.10948/) (2025)
- [Is Large-Scale Pretraining the Secret to Good Domain Generalization?](https://mlsanity.com/paper/2412.02856/) (2024)

## Built on by (most cited first, 2 of 2)

- [The ARC of Progress towards AGI: A Living Survey of Abstraction and Reasoning](https://mlsanity.com/paper/2603.13372/) (2026)
- [Beyond the Training Distribution: Mapping Generalization Boundaries in Neural Program Synthesis](https://mlsanity.com/paper/2604.27551/) (2026)

## Abstract (arXiv)

We run a controlled compositional generalization experiment in the ARC-AGI domain: an open-world problem domain in which the ability to generalize out-of-distribution is, by design, an essential characteristic for success. We compare neural program synthesis and test-time fine-tuning approaches on this experiment. We find that execution-guided neural program synthesis outperforms all reference algorithms in its ability to compose novel solutions. Our empirical findings also suggest that the success of TTFT on ARC-AGI lies mainly in eliciting in-distribution knowledge that the LLM otherwise fails to rely on directly.