PVF (Parameter Vulnerability Factor): A Scalable Metric for Understanding AI Vulnerability Against SDCs in Model Parameters

• URL: http://arxiv.org/abs/2405.01741v3
• Date: Tue, 11 Jun 2024 22:37:33 GMT
• Title: PVF (Parameter Vulnerability Factor): A Scalable Metric for Understanding AI Vulnerability Against SDCs in Model Parameters
• Authors: Xun Jiao, Fred Lin, Harish D. Dixit, Joel Coburn, Abhinav Pandey, Han Wang, Venkat Ramesh, Jianyu Huang, Wang Xu, Daniel Moore, Sriram Sankar,
• Abstract summary: Vulnerability Factor (PVF) is a metric aiming to standardize the quantification of AI model vulnerability against parameter corruptions. PVF can provide pivotal insights to AI hardware designers in balancing the tradeoff between fault protection and performance/efficiency. We present several use cases on applying PVF to three types of tasks/models during inference -- recommendation (DLRM), vision classification (CNN), and text classification (BERT)
• Score: 7.652441604508354
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Reliability of AI systems is a fundamental concern for the successful deployment and widespread adoption of AI technologies. Unfortunately, the escalating complexity and heterogeneity of AI hardware systems make them increasingly susceptible to hardware faults, e.g., silent data corruptions (SDC), that can potentially corrupt model parameters. When this occurs during AI inference/servicing, it can potentially lead to incorrect or degraded model output for users, ultimately affecting the quality and reliability of AI services. In light of the escalating threat, it is crucial to address key questions: How vulnerable are AI models to parameter corruptions, and how do different components (such as modules, layers) of the models exhibit varying vulnerabilities to parameter corruptions? To systematically address this question, we propose a novel quantitative metric, Parameter Vulnerability Factor (PVF), inspired by architectural vulnerability factor (AVF) in computer architecture community, aiming to standardize the quantification of AI model vulnerability against parameter corruptions. We define a model parameter's PVF as the probability that a corruption in that particular model parameter will result in an incorrect output. In this paper, we present several use cases on applying PVF to three types of tasks/models during inference -- recommendation (DLRM), vision classification (CNN), and text classification (BERT), while presenting an in-depth vulnerability analysis on DLRM. PVF can provide pivotal insights to AI hardware designers in balancing the tradeoff between fault protection and performance/efficiency such as mapping vulnerable AI parameter components to well-protected hardware modules. PVF metric is applicable to any AI model and has a potential to help unify and standardize AI vulnerability/resilience evaluation practice.

Related papers

• Enhanced Model Robustness to Input Corruptions by Per-corruption Adaptation of Normalization Statistics [22.876222327262596]
  We introduce Per-corruption Adaptation of Normalization statistics (PAN) to enhance the model robustness of vision systems. Our approach entails three key components: (i) a corruption type identification module, (ii) dynamic adjustment of normalization layer statistics based on identified corruption type, and (iii) real-time update of these statistics according to input data.
  arXiv  Detail & Related papers  (2024-07-08T23:20:18Z)
• AI Sandbagging: Language Models can Strategically Underperform on Evaluations [1.0485739694839669]
  Trustlocked AI systems are crucial for ensuring the safety of AI systems. Developers of AI systems may have incentives for sandbagging evaluations. We show that capability evaluations are vulnerable to sandbagging.
  arXiv  Detail & Related papers  (2024-06-11T15:26:57Z)
• Analyzing Adversarial Inputs in Deep Reinforcement Learning [53.3760591018817]
  We present a comprehensive analysis of the characterization of adversarial inputs, through the lens of formal verification. We introduce a novel metric, the Adversarial Rate, to classify models based on their susceptibility to such perturbations. Our analysis empirically demonstrates how adversarial inputs can affect the safety of a given DRL system with respect to such perturbations.
  arXiv  Detail & Related papers  (2024-02-07T21:58:40Z)
• STEAM & MoSAFE: SOTIF Error-and-Failure Model & Analysis for AI-Enabled Driving Automation [4.820785104084241]
  This paper defines the SOTIF Temporal Error and Failure Model (STEAM) as a refinement of the SOTIF cause-and-effect model. Second, this paper proposes the Model-based SOTIF Analysis of Failures and Errors (MoSAFE) method, which allows instantiating STEAM based on system-design models.
  arXiv  Detail & Related papers  (2023-12-15T06:34:35Z)
• Intrusion Detection System with Machine Learning and Multiple Datasets [0.0]
  In this paper, an enhanced intrusion detection system (IDS) that utilizes machine learning (ML) is explored. Ultimately, this improved system can be used to combat the attacks made by unethical hackers.
  arXiv  Detail & Related papers  (2023-12-04T14:58:19Z)
• ASSERT: Automated Safety Scenario Red Teaming for Evaluating the Robustness of Large Language Models [65.79770974145983]
  ASSERT, Automated Safety Scenario Red Teaming, consists of three methods -- semantically aligned augmentation, target bootstrapping, and adversarial knowledge injection. We partition our prompts into four safety domains for a fine-grained analysis of how the domain affects model performance. We find statistically significant performance differences of up to 11% in absolute classification accuracy among semantically related scenarios and error rates of up to 19% absolute error in zero-shot adversarial settings.
  arXiv  Detail & Related papers  (2023-10-14T17:10:28Z)
• HuntGPT: Integrating Machine Learning-Based Anomaly Detection and Explainable AI with Large Language Models (LLMs) [0.09208007322096533]
  We present HuntGPT, a specialized intrusion detection dashboard applying a Random Forest classifier. The paper delves into the system's architecture, components, and technical accuracy, assessed through Certified Information Security Manager (CISM) Practice Exams. The results demonstrate that conversational agents, supported by LLM and integrated with XAI, provide robust, explainable, and actionable AI solutions in intrusion detection.
  arXiv  Detail & Related papers  (2023-09-27T20:58:13Z)
• Guiding AI-Generated Digital Content with Wireless Perception [69.51950037942518]
  We introduce an integration of wireless perception with AI-generated content (AIGC) to improve the quality of digital content production. The framework employs a novel multi-scale perception technology to read user's posture, which is difficult to describe accurately in words, and transmits it to the AIGC model as skeleton images. Since the production process imposes the user's posture as a constraint on the AIGC model, it makes the generated content more aligned with the user's requirements.
  arXiv  Detail & Related papers  (2023-03-26T04:39:03Z)
• Enhancing Multiple Reliability Measures via Nuisance-extended Information Bottleneck [77.37409441129995]
  In practical scenarios where training data is limited, many predictive signals in the data can be rather from some biases in data acquisition. We consider an adversarial threat model under a mutual information constraint to cover a wider class of perturbations in training. We propose an autoencoder-based training to implement the objective, as well as practical encoder designs to facilitate the proposed hybrid discriminative-generative training.
  arXiv  Detail & Related papers  (2023-03-24T16:03:21Z)
• Federated Learning with Unreliable Clients: Performance Analysis and Mechanism Design [76.29738151117583]
  Federated Learning (FL) has become a promising tool for training effective machine learning models among distributed clients. However, low quality models could be uploaded to the aggregator server by unreliable clients, leading to a degradation or even a collapse of training. We model these unreliable behaviors of clients and propose a defensive mechanism to mitigate such a security risk.
  arXiv  Detail & Related papers  (2021-05-10T08:02:27Z)
• Trustworthy AI [75.99046162669997]
  Brittleness to minor adversarial changes in the input data, ability to explain the decisions, address the bias in their training data, are some of the most prominent limitations. We propose the tutorial on Trustworthy AI to address six critical issues in enhancing user and public trust in AI systems.
  arXiv  Detail & Related papers  (2020-11-02T20:04:18Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.