2019/12/19 by Stefano Teso, Teso, Stefano
Computer Science · Mathematics · #Adversarial Robustness in Machine Learning #Bayesian Modeling and Causal Inference #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #cs.LG #stat.ML
paper · pdf · doi:10.48550/arxiv.1912.10834
Short position paper. Accepted at the Ninth International Workshop on Statistical Relational AI (StarIA 2020)
arxiv created 2019/12/19 · openalex publication_date 2019/12/19 · arxiv updated 2019/12/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Arguments in favor of injecting symbolic knowledge into neural architectures abound. When done right, constraining a sub-symbolic model can substantially improve its performance and sample complexity and prevent it from predicting invalid configurations. Focusing on deep probabilistic (logical) graphical models -- i.e., constrained joint distributions whose parameters are determined (in part) by neural nets based on low-level inputs -- we draw attention to an elementary but unintended consequence of symbolic knowledge: that the resulting constraints can propagate the negative effects of adversarial examples.