arXiv subjects

Hana Chockler

Publications and source records attributed to Hana Chockler.

45 records · Page 3Linked to original sources

arXiv2021

Ranking Policy Decisions

Policies trained via Reinforcement Learning (RL) are often needlessly complex, making them difficult to analyse and interpret. In a run with $n$ time steps, a policy will make $n$ decisions on actions to take; we conjecture that only a small subset of these decisions delivers value over selecting a simple default action. Given a trained policy, we propose a novel black-box method based on statistical fault localisation that ranks the states of the environment according to the importance of decisions made in those states. We argue that among other things, the ranked list of states can help explain and understand the policy. As the ranking method is statistical, a direct evaluation of its quality is hard. As a proxy for quality, we use the ranking to create new, simpler policies from the original ones by pruning decisions identified as unimportant (that is, replacing them by default actions) and measuring the impact on performance. Our experiments on a diverse set of standard benchmarks demonstrate that pruned policies can perform on a level comparable to the original policies. Conversely, we show that naive approaches for ranking policy decisions, e.g., ranking based on the frequency of visiting a state, do not result in high-performing pruned policies.

cs.LG↗

arXiv2021

Explanations for Occluded Images

Existing algorithms for explaining the output of image classifiers perform poorly on inputs where the object of interest is partially occluded. We present a novel, black-box algorithm for computing explanations that uses a principled approach based on causal theory. We have implemented the method in the DEEPCOVER tool. We obtain explanations that are much more accurate than those generated by the existing explanation tools on images with occlusions and observe a level of performance comparable to the state of the art when explaining images without occlusions.

cs.LG↗

arXiv2020

Safety Synthesis Sans Specification

We define the problem of learning a transducer ${S}$ from a target language $U$ containing possibly conflicting transducers, using membership queries and conjecture queries. The requirement is that the language of ${S}$ be a subset of $U$. We argue that this is a natural question in many situations in hardware and software verification. We devise a learning algorithm for this problem and show that its time and query complexity is polynomial with respect to the rank of the target language, its incompatibility measure, and the maximal length of a given counterexample. We report on experiments conducted with a prototype implementation.

cs.FL↗

arXiv2020

Explaining Image Classifiers using Statistical Fault Localization

The black-box nature of deep neural networks (DNNs) makes it impossible to understand why a particular output is produced, creating demand for "Explainable AI". In this paper, we show that statistical fault localization (SFL) techniques from software engineering deliver high quality explanations of the outputs of DNNs, where we define an explanation as a minimal subset of features sufficient for making the same decision as for the original input. We present an algorithm and a tool called DeepCover, which synthesizes a ranking of the features of the inputs using SFL and constructs explanations for the decisions of the DNN based on this ranking. We compare explanations produced by DeepCover with those of the state-of-the-art tools GradCAM, LIME, SHAP, RISE and Extremal and show that explanations generated by DeepCover are consistently better across a broad set of experiments. On a benchmark set with known ground truth, DeepCover achieves 76.7% accuracy, which is 6% better than the second best Extremal.

cs.LG↗

arXiv2020

Combining Experts' Causal Judgments

Consider a policymaker who wants to decide which intervention to perform in order to change a currently undesirable situation. The policymaker has at her disposal a team of experts, each with their own understanding of the causal dependencies between different factors contributing to the outcome. The policymaker has varying degrees of confidence in the experts' opinions. She wants to combine their opinions in order to decide on the most effective intervention. We formally define the notion of an effective intervention, and then consider how experts' causal judgments can be combined in order to determine the most effective intervention. We define a notion of two causal models being \emph{compatible}, and show how compatible causal models can be merged. We then use it as the basis for combining experts' causal judgments. We also provide a definition of decomposition for causal models to cater for cases when models are incompatible. We illustrate our approach on a number of real-life examples.

cs.AI↗

arXiv2016

Causality and Responsibility for Formal Verification and Beyond

The theory of actual causality, defined by Halpern and Pearl, and its quantitative measure - the degree of responsibility - was shown to be extremely useful in various areas of computer science due to a good match between the results it produces and our intuition. In this paper, I describe the applications of causality to formal verification, namely, explanation of counterexamples, refinement of coverage metrics, and symbolic trajectory evaluation. I also briefly discuss recent applications of causality to legal reasoning.

cs.SE↗

arXiv2014

The Computational Complexity of Structure-Based Causality

Halpern and Pearl introduced a definition of actual causality; Eiter and Lukasiewicz showed that computing whether X=x is a cause of Y=y is NP-complete in binary models (where all variables can take on only two values) and\ Sigma_2^P-complete in general models. In the final version of their paper, Halpern and Pearl slightly modified the definition of actual cause, in order to deal with problems pointed by Hopkins and Pearl. As we show, this modification has a nontrivial impact on the complexity of computing actual cause. To characterize the complexity, a new family D_k^P, k= 1, 2, 3, ..., of complexity classes is introduced, which generalizes the class DP introduced by Papadimitriou and Yannakakis (DP is just D_1^P). %joe2 %We show that the complexity of computing causality is $\D_2$-complete %under the new definition. Chockler and Halpern \citeyear{CH04} extended the We show that the complexity of computing causality under the updated definition is $D_2^P$-complete. Chockler and Halpern extended the definition of causality by introducing notions of responsibility and blame. The complexity of determining the degree of responsibility and blame using the original definition of causality was completely characterized. Again, we show that changing the definition of causality affects the complexity, and completely characterize it using the updated definition.

cs.AI↗

arXiv2003

What Causes a System to Satisfy a Specification?

Even when a system is proven to be correct with respect to a specification, there is still a question of how complete the specification is, and whether it really covers all the behaviors of the system. Coverage metrics attempt to check which parts of a system are actually relevant for the verification process to succeed. Recent work on coverage in model checking suggests several coverage metrics and algorithms for finding parts of the system that are not covered by the specification. The work has already proven to be effective in practice, detecting design errors that escape early verification efforts in industrial settings. In this paper, we relate a formal definition of causality given by Halpern and Pearl [2001] to coverage. We show that it gives significant insight into unresolved issues regarding the definition of coverage and leads to potentially useful extensions of coverage. In particular, we introduce the notion of responsibility, which assigns to components of a system a quantitative measure of their relevance to the satisfaction of the specification.

cs.LO↗

arXiv2003

Responsibility and blame: a structural-model approach

Causality is typically treated an all-or-nothing concept; either A is a cause of B or it is not. We extend the definition of causality introduced by Halpern and Pearl [2001] to take into account the degree of responsibility of A for B. For example, if someone wins an election 11--0, then each person who votes for him is less responsible for the victory than if he had won 6--5. We then define a notion of degree of blame, which takes into account an agent's epistemic state. Roughly speaking, the degree of blame of A for B is the expected degree of responsibility of A for B, taken over the epistemic state of an agent.

cs.AI↗