Probably Approximately Correct Explanations of Machine Learning Models via Syntax-Guided Synthesis

Neider, Daniel; Ghosh, Bishwamittra

doi:10.48550/arxiv.2009.08770

Cited by 2 publications

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our approach for computing formal explanations from examples follows this idea. Neider et al [32] followed a similar direction and proposed to use a combination of probably approximately correct learning (PAC) and syntax-guided synthesis (SyGuS) to produce explanations that with a high probability make only few errors. In contrast to our work, [32] does not compute explanations for image data.…”

Section: Related Workmentioning

confidence: 99%

Formal XAI via Syntax-Guided Synthesis

Bjørner,

Judson,

Cano

et al. 2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

In this paper, we propose a novel application of syntaxguided synthesis to find symbolic representations of a model's decisionmaking process, designed for easy comprehension and validation by humans. Our approach takes input-output samples from complex machine learning models, such as deep neural networks, and automatically derives interpretable mimic programs. A mimic program precisely imitates the behavior of an opaque model over the provided data. We discuss various types of grammars that are well-suited for computing mimic programs for tabular and image input data. Our experiments demonstrate the potential of the proposed method: we successfully synthesized mimic programs for neural networks trained on the MNIST and the Pima Indians diabetes data sets. All experiments were performed using the SMT-based cvc5 synthesis tool.

show abstract

Section: Related Workmentioning

confidence: 99%

Formal XAI via Syntax-Guided Synthesis

Bjørner,

Judson,

Cano

et al. 2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

Mining Hyperproperties using Temporal Logics

Bartocci,

Mateis,

Nesterini

et al. 2023

ACM Trans. Embed. Comput. Syst.

View full text Add to dashboard Cite

Formal specifications are essential to express precisely systems, but they are often difficult to define or unavailable. Specification mining aims to automatically infer specifications from system executions. The existing literature mainly focuses on learning properties defined on single system executions. However, many system characteristics, such as security policies and robustness, require relating two or more executions, and hence cannot be captured by properties. Hyperproperties address this limitation by allowing simultaneous reasoning about multiple executions with quantification over system traces. In this paper, we propose an effective approach for mining Hyper Signal Temporal Logic (HyperSTL) specifications. Our approach is based on the syntax-guided synthesis framework and allows users to control the amount of prior knowledge embedded in the mining procedure. To the best of our knowledge, this is the first mining method for hyperproperties that does not require a pre-defined template as input and allows for quantifier alternation. We implemented our approach and demonstrated its applicability and versatility in several case studies where we showed that we can use the same method to mine specifications both with and without templates, but also to infer subsets of HyperSTL, including STL, HyperLTL, LTL and non-temporal specifications.

show abstract

Probably Approximately Correct Explanations of Machine Learning Models via Syntax-Guided Synthesis

Cited by 2 publications

References 10 publications

Formal XAI via Syntax-Guided Synthesis

Formal XAI via Syntax-Guided Synthesis

Mining Hyperproperties using Temporal Logics

Contact Info

Product

Resources

About