Export Citations
Save this search
Please login to be able to save your searches and receive alerts for new content matching your search criteria.
- rapid-communicationMay 2024
Chordal sparsity for SDP-based neural network verification
Automatica (Journal of IFAC) (AJIF), Volume 161, Issue Chttps://doi.org/10.1016/j.automatica.2023.111487AbstractNeural networks are central to many emerging technologies, but verifying their correctness remains a major challenge. It is known that network outputs can be sensitive and fragile to even small input perturbations, thereby increasing the risk of ...
- research-articleMay 2024
Stability guarantees for feature attributions with multiplicative smoothing
NIPS '23: Proceedings of the 37th International Conference on Neural Information Processing SystemsArticle No.: 2724, Pages 62388–62413Explanation methods for machine learning models tend not to provide any formal guarantees and may not reflect the underlying decision-making process. In this work, we analyze stability as a property for reliable feature attribution methods. We prove that ...
- research-articleOctober 2023
Mobius: Synthesizing Relational Queries with Recursive and Invented Predicates
Proceedings of the ACM on Programming Languages (PACMPL), Volume 7, Issue OOPSLA2Article No.: 271, Pages 1394–1417https://doi.org/10.1145/3622847Synthesizing relational queries from data is challenging in the presence of recursion and invented predicates. We propose a fully automated approach to synthesize such queries. Our approach comprises of two steps: it first synthesizes a non-recursive ...
Relational Query Synthesis ⋈ Decision Tree Learning
Proceedings of the VLDB Endowment (PVLDB), Volume 17, Issue 2Pages 250–263https://doi.org/10.14778/3626292.3626306We study the problem of synthesizing a core fragment of relational queries called select-project-join (SPJ) queries from input-output examples. Search-based synthesis techniques are suited to synthesizing projections and joins by navigating the network ...
- research-articleJuly 2023
Robust subtask learning for compositional generalization
ICML'23: Proceedings of the 40th International Conference on Machine LearningArticle No.: 627, Pages 15371–15387Compositional reinforcement learning is a promising approach for training policies to perform complex long-horizon tasks. Typically, a high-level task is decomposed into a sequence of subtasks and a separate policy is trained to perform each subtask. In ...
-
- ArticleJuly 2023
Policy Synthesis and Reinforcement Learning for Discounted LTL
AbstractThe difficulty of manually specifying reward functions has led to an interest in using linear temporal logic (LTL) to express objectives for reinforcement learning (RL). However, LTL has the downside that it is sensitive to small perturbations in ...
- research-articleJanuary 2023
A Robust Theory of Series Parallel Graphs
Proceedings of the ACM on Programming Languages (PACMPL), Volume 7, Issue POPLArticle No.: 37, Pages 1058–1088https://doi.org/10.1145/3571230Motivated by distributed data processing applications, we introduce a class of labeled directed acyclic graphs constructed using sequential and parallel composition operations, and study automata and logics over them. We show that deterministic and non-...
Executing Microservice Applications on Serverless, Correctly
Proceedings of the ACM on Programming Languages (PACMPL), Volume 7, Issue POPLArticle No.: 13, Pages 367–395https://doi.org/10.1145/3571206While serverless platforms substantially simplify the provisioning, configuration, and management of cloud applications, implementing correct services on top of these platforms can present significant challenges to programmers. For example, serverless ...
- ArticleApril 2022
Automatic Repair for Network Programs
Tools and Algorithms for the Construction and Analysis of SystemsPages 353–372https://doi.org/10.1007/978-3-030-99527-0_19AbstractDebugging imperative network programs is a difficult task for operators as it requires understanding various network modules and complicated data structures. For this purpose, this paper presents an automated technique for repairing network ...
Stream processing with dependency-guided synchronization
PPoPP '22: Proceedings of the 27th ACM SIGPLAN Symposium on Principles and Practice of Parallel ProgrammingPages 1–16https://doi.org/10.1145/3503221.3508413Real-time data processing applications with low latency requirements have led to the increasing popularity of stream processing systems. While such systems offer convenient APIs that can be used to achieve data parallelism automatically, they offer ...
- chapterMarch 2022
Continuous-Time Models for System Design and Analysis
AbstractWe illustrate the ingredients of the state-of-the-art of model-based approach for the formal design and verification of cyber-physical systems. To capture the interaction between a discrete controller and its continuously evolving environment, we ...
- research-articleJune 2024
Compositional reinforcement learning from logical specifications
NIPS '21: Proceedings of the 35th International Conference on Neural Information Processing SystemsArticle No.: 767, Pages 10026–10039We study the problem of learning control policies for complex tasks given by logical specifications. Recent approaches automatically generate a reward function from a given specification and use a suitable reinforcement learning algorithm to learn a ...
- research-articleNovember 2021
- research-articleSeptember 2021
Compositional Learning and Verification of Neural Network Controllers
ACM Transactions on Embedded Computing Systems (TECS), Volume 20, Issue 5sArticle No.: 92, Pages 1–26https://doi.org/10.1145/3477023Recent advances in deep learning have enabled data-driven controller design for autonomous systems. However, verifying safety of such controllers, which are often hard-to-analyze neural networks, remains a challenge. Inspired by compositional strategies ...
- ArticleJuly 2021
Verisig 2.0: Verification of Neural Network Controllers Using Taylor Model Preconditioning
AbstractThis paper presents Verisig 2.0, a verification tool for closed-loop systems with neural network (NN) controllers. We focus on NNs with tanh/sigmoid activations and develop a Taylor-model-based reachability algorithm through Taylor model ...
- research-articleJune 2021
Synchronization Schemas
- Rajeev Alur,
- Phillip Hilliard,
- Zachary G. Ives,
- Konstantinos Kallas,
- Konstantinos Mamouras,
- Filip Niksic,
- Caleb Stanford,
- Val Tannen,
- Anton Xue
PODS'21: Proceedings of the 40th ACM SIGMOD-SIGACT-SIGAI Symposium on Principles of Database SystemsPages 1–18https://doi.org/10.1145/3452021.3458317We present a type-theoretic framework for data stream processing for real-time decision making, where the desired computation involves a mix of sequential computation, such as smoothing and detection of peaks and surges, and naturally parallel ...
- research-articleJune 2021
Example-guided synthesis of relational queries
PLDI 2021: Proceedings of the 42nd ACM SIGPLAN International Conference on Programming Language Design and ImplementationPages 1110–1125https://doi.org/10.1145/3453483.3454098Program synthesis tasks are commonly specified via input-output examples. Existing enumerative techniques for such tasks are primarily guided by program syntax and only make indirect use of the examples. We identify a class of synthesis algorithms for ...
- ArticleMarch 2021
Network Traffic Classification by Program Synthesis
Tools and Algorithms for the Construction and Analysis of SystemsPages 430–448https://doi.org/10.1007/978-3-030-72016-2_23AbstractWriting classification rules to identify interesting network traffic is a time-consuming and error-prone task. Learning-based classification systems automatically extract such rules from positive and negative traffic examples. However, due to ...
- research-articleMarch 2021
Static detection of uncoalesced accesses in GPU programs
Formal Methods in System Design (FMSD), Volume 60, Issue 1Pages 1–32https://doi.org/10.1007/s10703-021-00362-8AbstractGPU programming has become popular due to the high computational capabilities of GPUs. Obtaining significant performance gains with GPU is however challenging and the programmer needs to be aware of various subtleties of the GPU architecture. One ...