Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2510.13163 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917015414898688 |
|---|---|
| author | Iskandar, Nyx Bedri, Hisham Tsen, Andy |
| author_facet | Iskandar, Nyx Bedri, Hisham Tsen, Andy |
| contents | Most large language models (LLMs) today excel at generating raw, sequential code with minimal abstractions and custom structures. However, there has been little work on graph-based abstract code generation, where significant logic is encapsulated in predefined nodes and execution flow is determined by edges. This is relevant for visual programming languages, and in cases where raw source code is inaccessible to users and LLM training sets. In this work, we propose and evaluate JSON representations for graphs to enable high accuracy graph-based abstract code generation. We evaluate these representations on ScratchTest, a mini-benchmark based on our custom Python re-implementation of Scratch, which tests the LLM in code graph space. Our findings demonstrate that LLMs can indeed perform the aforementioned generation task in a single pass without relying on specialized or complex pipelines, given the correct graph representations. We also show that different representations induce significantly different accuracies, highlighting the instrumental role of representations in this generation task. All in all, this work establishes the first steps towards representation learning for graph-based abstract code generation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_13163 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A Matter of Representation: Towards Graph-Based Abstract Code Generation Iskandar, Nyx Bedri, Hisham Tsen, Andy Computation and Language Most large language models (LLMs) today excel at generating raw, sequential code with minimal abstractions and custom structures. However, there has been little work on graph-based abstract code generation, where significant logic is encapsulated in predefined nodes and execution flow is determined by edges. This is relevant for visual programming languages, and in cases where raw source code is inaccessible to users and LLM training sets. In this work, we propose and evaluate JSON representations for graphs to enable high accuracy graph-based abstract code generation. We evaluate these representations on ScratchTest, a mini-benchmark based on our custom Python re-implementation of Scratch, which tests the LLM in code graph space. Our findings demonstrate that LLMs can indeed perform the aforementioned generation task in a single pass without relying on specialized or complex pipelines, given the correct graph representations. We also show that different representations induce significantly different accuracies, highlighting the instrumental role of representations in this generation task. All in all, this work establishes the first steps towards representation learning for graph-based abstract code generation. |
| title | A Matter of Representation: Towards Graph-Based Abstract Code Generation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2510.13163 |