Enabling Global, Human-Centered Explanations for LLMs:From Tokens to Interpretable Code and Test Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915932554657792 |
|---|---|
| author | Khati, Dipin Rodriguez-Cardenas, Daniel Palacio, David N. Velasco, Alejandro Tufano, Michele Poshyvanyk, Denys |
| author_facet | Khati, Dipin Rodriguez-Cardenas, Daniel Palacio, David N. Velasco, Alejandro Tufano, Michele Poshyvanyk, Denys |
| contents | As Large Language Models for Code (LM4Code) become integral to software engineering, establishing trust in their output becomes critical. However, standard accuracy metrics obscure the underlying reasoning of generative models, offering little insight into how decisions are made. Although post-hoc interpretability methods attempt to fill this gap, they often restrict explanations to local, token-level insights, which fail to provide a developer-understandable global analysis. Our work highlights the urgent need for \textbf{global, code-based} explanations that reveal how models reason across code. To support this vision, we introduce \textit{code rationales} (CodeQ), a framework that enables global interpretability by mapping token-level rationales to high-level programming categories. Aggregating thousands of these token-level explanations allows us to perform statistical analyses that expose systemic reasoning behaviors. We validate this aggregation by showing it distills a clear signal from noisy token data, reducing explanation uncertainty (Shannon entropy) by over 50%. Additionally, we find that a code generation model (\textit{codeparrot-small}) consistently favors shallow syntactic cues (e.g., \textbf{indentation}) over deeper semantic logic. Furthermore, in a user study with 37 participants, we find its reasoning is significantly misaligned with that of human developers. These findings, hidden from traditional metrics, demonstrate the importance of global interpretability techniques to foster trust in LM4Code. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_16771 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Enabling Global, Human-Centered Explanations for LLMs:From Tokens to Interpretable Code and Test Generation Khati, Dipin Rodriguez-Cardenas, Daniel Palacio, David N. Velasco, Alejandro Tufano, Michele Poshyvanyk, Denys Software Engineering Machine Learning As Large Language Models for Code (LM4Code) become integral to software engineering, establishing trust in their output becomes critical. However, standard accuracy metrics obscure the underlying reasoning of generative models, offering little insight into how decisions are made. Although post-hoc interpretability methods attempt to fill this gap, they often restrict explanations to local, token-level insights, which fail to provide a developer-understandable global analysis. Our work highlights the urgent need for \textbf{global, code-based} explanations that reveal how models reason across code. To support this vision, we introduce \textit{code rationales} (CodeQ), a framework that enables global interpretability by mapping token-level rationales to high-level programming categories. Aggregating thousands of these token-level explanations allows us to perform statistical analyses that expose systemic reasoning behaviors. We validate this aggregation by showing it distills a clear signal from noisy token data, reducing explanation uncertainty (Shannon entropy) by over 50%. Additionally, we find that a code generation model (\textit{codeparrot-small}) consistently favors shallow syntactic cues (e.g., \textbf{indentation}) over deeper semantic logic. Furthermore, in a user study with 37 participants, we find its reasoning is significantly misaligned with that of human developers. These findings, hidden from traditional metrics, demonstrate the importance of global interpretability techniques to foster trust in LM4Code. |
| title | Enabling Global, Human-Centered Explanations for LLMs:From Tokens to Interpretable Code and Test Generation |
| topic | Software Engineering Machine Learning |
| url | https://arxiv.org/abs/2503.16771 |