论形式化反事实解释
On Formal Counterfactual Explanations
AI 导读
一篇 arXiv 论文提出将最接近世界语义的反事实解释显式刻画为所有最小距离反例实例上的析取范式,并引入合取反事实解释(CFXps)这一受限子类。论文建立了 CXps、AXps 与 CFXps 之间的结构关系:每个 CXp 可为其见证实例诱导一个 AXp,CXp 结合见证 AXp 可诱导一个 CFXp,从而复用现有 AXp/CXp 工具模块化计算 CFXps。
正文 · 原文
Abstract:Counterfactual explanations are often seen as counterfactual conditionals under closest-world semantics, requiring one to reason over all minimally distant counterexamples. In contrast, formal XAI typically relies on feature-based abductive explanations (AXps) and contrastive explanations (CXps), which admit efficient computation and a minimal hitting-set duality. This paper bridges these perspectives by (i) characterizing closest-world counterfactual explanations explicitly as a disjunctive normal form over all minimally distant counterexample instances, and (ii) introducing conjunctive counterfactual explanations (CFXps), a subclass of counterfactual antecedents restricted to conjunctions of feature literals that guarantee prediction change. The paper establishes structural results connecting CXps, AXps and CFXps: every CXp induces an AXp for its witness instance, and every CXp together with a witness AXp induces a CFXp whose support is precisely determined by that AXp. These theoretical results yield a modular approach for computing CFXps using existing AXp/CXp machinery. Experiments on a representative set of benchmarks using tree ensembles and neural networks show that CFXp extraction is practically efficient. Compared with popular point-based counterfactual explainers, CFXps generally require no more feature changes and can guarantee the counterfactual prediction beyond a single instance.
| Subjects: | Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Logic in Computer Science (cs.LO) |
| Cite as: | arXiv:2611.22010 [cs.AI] |
| (or arXiv:2611.22010v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.22010 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Alexey Ignatiev [view email]
[v1]
Mon, 16 Nov 2026 23:51:26 UTC (1,256 KB)
来源:arXiv cs.AI · arxiv.org