paper-with-me

홈 › Papers

Criticality and Safety Margins for Reinforcement Learning

2024-09-26 · Alexander Grushin, Walt Woods, Alvaro Velasquez, Simon Khan

State of the art reinforcement learning methods sometimes encounter unsafe situations. Identifying when these situations occur is of interest both for post-hoc analysis and during deployment, where it might be advantageous to call out to a human overseer for help. Efforts to gauge the criticality of different points in time have been developed, but their accuracy is not well established due to a lack of ground truth, and they are not designed to be easily interpretable by end users. Therefore, we seek to define a criticality framework with both a quantifiable ground truth and a clear significance to users. We introduce true criticality as the expected drop in reward when an agent deviates from its policy for n consecutive random actions. We also introduce the concept of proxy criticality, a low-overhead metric that has a statistically monotonic relationship to true criticality. Safety margins make these interpretable, when defined as the number of random actions for which performance loss will not exceed some tolerance with high confidence. We demonstrate this approach in several environment-agent combinations; for an A3C agent in an Atari Beamrider environment, the lowest 5% of safety margins contain 47% of agent losses; i.e., supervising only 5% of decisions could potentially prevent roughly half of an agent's errors. This criticality framework measures the potential impacts of bad decisions, even before those decisions are made, allowing for more effective debugging and oversight of autonomous agents.

📄 PDF Abstract BibTeX arXiv:2409.18289

Code (0)

등록된 구현이 없습니다.

Tasks

reinforcement-learningReinforcement Learning

Methods 이 논문이 사용한 방법론

Convolution A convolution is a type of matrix operation, consisting of a kernel, a small matrix of weights, that slides over input data performing element-wise multiplication with the…
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Entropy Regularization 설명 없음
A3C A3C, Asynchronous Advantage Actor Critic, is a policy gradient algorithm in reinforcement learning that maintains a policy $\pi\left(a\_{t}\mid{s}\_{t}; \theta\right)$ and…

Similar Papers 제목 키워드 기반

Safety Margins for Reinforcement Learning

2023-07-25 · Alexander Grushin, Walt Woods, Alvaro Velasquez, Simon Khan

Any autonomous controller will be unsafe in some situations. The ability to quantitatively identify when these unsafe situations are about to occur is crucial for drawing timely human oversight in, e.g., freight transpor…

reinforcement-learningReinforcement Learning

Improving Mixed-Criticality Scheduling with Reinforcement Learning

2025-04-04 · Muhammad El-Mahdy, Nourhan Sakr, Rodrigo Carrasco

This paper introduces a novel reinforcement learning (RL) approach to scheduling mixed-criticality (MC) systems on processors with varying speeds. Building upon the foundation laid by [1], we extend their work to address…

reinforcement-learningReinforcement LearningReinforcement Learning (RL)Scheduling

Criticality Metrics for Relevance Classification in Safety Evaluation of Object Detection in Automated Driving

2025-12-17 · Jörg Gamerdinger, Sven Teufel, Stephan Amann, Oliver Bringmann arxiv

Ensuring safety is the primary objective of automated driving, which necessitates a comprehensive and accurate perception of the environment. While numerous performance evaluation metrics exist for assessing perception c…

Object Detection

Accurately Predicting Probabilities of Safety-Critical Rare Events for Intelligent Systems

2024-03-20 · Ruoxuan Bai, Jingxuan Yang, Weiduo Gong, Yi Zhang 외

Intelligent systems are increasingly integral to our daily lives, yet rare safety-critical events present significant latent threats to their practical deployment. Addressing this challenge hinges on accurately predictin…

Multi-Agent Optimization for Safety Analysis of Cyber-Physical Systems: Position Paper

2024-03-25 · Önder Gürcan, Nataliya Yakymets, Sara Tucci-Piergiovanni, Ansgar Radermacher

Failure Mode, Effects and Criticality Analysis (FMECA) is one of the safety analysis methods recommended by most of the international standards. The classical FMECA is made in a form of a table filled in either manually …

Decision MakingPosition