The Adaptive Stress Testing Formulation
By Mark Koren, Anthony Corso, Mykel J. Kochenderfer. 2020
11 citations recorded by the source corpus.
Validation is a key challenge in the search for safe autonomy. Simulations are often either too simple to provide robust validation, or too complex to tractably compute. Therefore, approximate validation methods are needed to tractably find failures without unsafe simplifications. This paper presents the theory behind one such black-box approach: adaptive stress testing (AST). We also provide three examples of validation problems formulated to work with AST.
Subfields: Computational Theory and Mathematics, Control and Systems Engineering, Statistics and Probability
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computational Theory and Mathematics
- Adrian Weller
- Yoshua Bengio
- Jack Clark
- Shagun Sodhani
- Pang Wei Koh
- Miles Brundage
- Gillian K. Hadfield
- Peter Henderson
In Control and Systems Engineering
- David Krueger
- Adrian Weller
- Yoshua Bengio
- Shagun Sodhani
- Peter Eckersley
- Miles Brundage
- Peter Henderson
- Dougal Maclaurin
In Statistics and Probability
- David Krueger
- Adrian Weller
- Yoshua Bengio
- Shagun Sodhani
- Pang Wei Koh
- Peter Eckersley
- Peter Henderson
- Logan Graham
Citations in the corpus
- Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims (2020, methodology): “For example, adaptive stress testing (AST) of an AI system allows users to find the most likely failure of a system for a given scenario using reinforcement learning [61], and is being used by to validate the next generation of aircraft collision avoidance software [62].”