๐ก๐ณ๐ฝ๐ฝ ( ๐ฐ๐๐ฟ ) \mathtt{xDNN^{(ASP)}} : Explanation Generation System for Deep Neural Networks powered by Answer Set Programming
Ly Ly Trieu Tran Cao Son ยท 2026
Read ๐ก๐ณ๐ฝ๐ฝ ( ๐ฐ๐๐ฟ ) \mathtt{xDNN^{(ASP)}} : Explanation Generation System for Deep Neural Networks powered by Answer Set Programming by Ly Ly Trieu Tran Cao Son (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-03847.txt.
Description
Ly Ly Trieu and Tran Cao Son review and build an explanation system for deep neural networks powered by answer set programming. Deep networks are black boxes, and explanation methods differ by scope and style: some summarize behavior across all classes, others explain a single prediction. Well-known statistical tools like SHAP rank feature importance, rule extraction rewrites network behavior as if-then rules, and counterfactuals show the smallest change that flips a decision.
Their system, xDNN(ASP), sits in that landscape with a logic-based approach to generating explanations. The paper surveys the alternatives to position its method, contrasting statistic-based and rule-based strategies. Logic-backed explanations have a special appeal in regulated settings: they can be checked, enumerated, and argued with, rather than merely visualized as heat maps over the input.
Catalog record
| CREATOR | Ly Ly Trieu Tran Cao Son |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-03847