The paper discusses safety assessment for AI systems, focusing on machine learning models.
problem Safety assessment of AI systems, especially machine learning models, in safety-related applications.
method Analyzed AI models as statistical models and proposed a new budget allocation for AI safety.
result Safety assessment of AI systems requires a new approach focusing on the model used, not just the system.
Researchers introduce a method to assess the safety of interpretable machine learning models.
problem Ensuring safety in machine learning models that are easy to understand.
method Introduce maximum deviation as an optimization problem to find the largest deviation from a safe reference model.
result Interpretability helps in assessing the safety of machine learning models.
Safety evaluation of self-driving technologies has been extensively studied. One recent approach uses Monte Carlo based evaluation to estimate the occurrence probabilities of safety-critical events as safety measures. These Monte Carlo samples are generated from stochastic input models constructed based on real-world d…
Shai is a 10B model for asset management tasks, outperforming baselines.
problem Improving performance in asset management tasks.
method Continuous pre-training and fine-tuning on asset management-specific data.
result Shai outperforms baseline models in asset management tasks.
The safety and resilience of fully autonomous vehicles (AVs) are of significant concern, as exemplified by several headline-making accidents. While AV development today involves verification, validation, and testing, end-to-end assessment of AV systems under accidental faults in realistic driving scenarios has been lar…
The use of machine learning (ML) is on the rise in many sectors of software development, and automotive software development is no different. In particular, Advanced Driver Assistance Systems (ADAS) and Automated Driving Systems (ADS) are two areas where ML plays a significant role. In automotive development, safety is…
This paper formalizes AI safety using hypothesis testing in GenAI.
problem Ensuring safety of generative AI tools that create realistic content.
method Formalization of computational safety through hypothesis testing and signal processing.
result Demonstrates how AI safety can be assessed quantitatively using mathematical frameworks.
A framework identifies worst-case decision points in safety-critical scenarios, improving risk assessment by 10 hours.
problem Identifying worst-case outcomes in safety-critical decision-making under uncertainty.
method Explicitly estimating distributions of expected return to identify dead-ends, tuning based on risk tolerance.
result Significantly improves risk assessment, providing indications 10 hours earlier and increasing detection by 20%.
Improves pre-trial risk assessments by making them safer without changing existing rules.
problem Improving pre-trial risk assessments while maintaining deterministic rules.
method Developed a maximin robust optimization approach to find a safer policy.
result Can safely improve certain components of the risk assessment instrument.
Unified approach detects traffic conflicts across various interactions.
problem Inconsistent detection of traffic conflicts across different interactions.
method Unified probabilistic approach decomposes conflicts into statistical learning tasks.
result Effective collision warnings across diverse datasets and environments.
By building on a recently introduced genetic-inspired attribute-based conceptual framework for safety risk analysis, we propose a novel methodology to compute construction univariate and bivariate construction safety risk at a situational level. Our fully data-driven approach provides construction practitioners and aca…
This paper tests LLMs in finance to assess ethical behavior.
problem Aligning AI with ethical and legal standards in finance.
method Prompted LLMs to simulate CEO behavior, analyzed with logistic regression.
result Significant heterogeneity in LLMs' unethical behavior propensity.
The prediction of workers' safety behaviour can help identify vulnerable workers who intend to undertake unsafe behaviours and be useful in the design of management practices to minimise the occurrence of accidents. The latest literature has evidenced that there is within-population diversity that leads people's intend…
The paper evaluates and improves uncertainty estimates in neural networks for safety-critical applications.
problem Quantifying uncertainty in neural networks for safety-critical systems.
method Proposes a statistical test for evaluating uncertainty realism in neural networks and transfers a classification architecture to image-to-image tasks.
result The variational U-Net architecture significantly improves uncertainty realism in image-to-image tasks compared to a plain model.
Paper explores physics-informed deep learning for system reliability assessment.
problem Limited study on deep learning for system reliability assessment.
method Physics-informed deep learning approach for system reliability assessment.
result Physics-informed deep learning can alleviate computational challenges and combine measurement data and mathematical models.
Quantum model improves safety in machine learning.
problem Improving safety and robustness in machine learning models.
method Variational quantum classifier with amplitude encoding and SAFE-AI metrics.
result Quantum model provides competitive performance and improved robustness.
Affective states have a critical role in driving performance and safety. They can degrade driver situation awareness and negatively impact cognitive processes, severely diminishing road safety. Therefore, detecting and assessing drivers' affective states is crucial in order to help improve the driving experience, and i…
AI systems need reliable testing to ensure safety and trustworthiness.
problem Current AI Act lacks functional trustworthiness for AI systems.
method Define technical application distribution, set risk-based performance, and conduct statistically valid testing.
result Reliable functional trustworthiness is essential for AI systems.
We study the risk assessment of uncertain cash flows in terms of dynamic convex risk measures for processes as introduced in Cheridito, Delbaen, and Kupper (2006). These risk measures take into account not only the amounts but also the timing of a cash flow. We discuss their robust representation in terms of suitably p…
The paper assesses how equity tail risk impacts US Treasury bond returns.
problem The effects of equity tail risk on the US government bond market.
method Estimating equity tail risk using option-implied stock market volatility and assessing its predictive power in reduced-form regressions and a term structure model.
result Equity tail risk significantly predicts one-month excess returns on Treasuries.
CPA framework assesses conditional validity of conformal prediction.
problem Challenges in evaluating conditional validity of conformal prediction.
method Reframes conditional coverage evaluation as a supervised learning task.
result Establishes convergence rates and proves CVI consistency.
Enhances crowd safety through AI and data-driven models.
problem Improving crowd safety during events.
method Innovative data collection, AI, and machine learning.
result Accurate multi-day forecasts for event planning.
We develop a method to estimate the time to unsafe responses in LLMs.
problem Estimating the time to unsafe responses in large language models is challenging due to the rarity of unsafe outputs.
method We frame the problem as survival analysis and propose a calibration technique for constructing a lower predictive bound (LPB).
result Our method provides rigorous coverage guarantees and improves sample efficiency.
Paper proposes model to assess financial risk of grid-ignited wildfires.
problem Financial risk and solvency threats from grid-ignited wildfires.
method Integrated model to evaluate damage costs and risk levels.
result Identifies high-risk areas for preemptive actions.
ContextBench benchmarks methods for generating linguistically fluent inputs that activate specific latent features in language models.
problem Identifying inputs that trigger specific behaviours or latent features in language models.
method Context modification and benchmarking methods like Evolutionary Prompt Optimisation (EPO) with LLM-assistance and diffusion model inpainting.
result Enhanced methods achieve state-of-the-art performance in balancing elicitation effectiveness and fluency.
Radio-frequency dosimetry is an important process in human safety and for compliance of related products. Recently, computational human models generated from medical images have often been used for such assessment, especially to consider the inter-variability of subjects. However, the common procedure to develop person…
The paper proposes a method to assess when automated predictions are reliable.
problem Ensuring reliability and safety of automated decision-making in machine learning.
method Clustering to measure distances between outputs and class centroids, defining a safety threshold based on these distances.
result The proposed metric can efficiently determine when automated predictions are acceptable and when they should be deferred.
Study assesses weakly-supervised methods for rare outcomes in medical records.
problem Identifying patients with specific medical conditions using electronic health records.
method Compared three methods (PheNorm, MAP, and sureLDA) in simulations with varying outcomes and silver labels.
result No single method consistently outperformed others, but sureLDA often did well.
Anomalies (unusual patterns) in time-series data give essential, and often actionable information in critical situations. Examples can be found in such fields as healthcare, intrusion detection, finance, security and flight safety. In this paper we propose new conformalized density- and distance-based anomaly detection…
Combines Lyapunov functions with controller synthesis for safe control policies.
problem Ensuring safety in controller design for nonlinear systems.
method Iterative algorithm combining Lyapunov function estimation and controller synthesis.
result Effective control policies with large safe regions are derived.
It has been shown that deep neural networks (DNNs) may be vulnerable to adversarial attacks, raising the concern on their robustness particularly for safety-critical applications. Recognizing the local nature and limitations of existing adversarial attacks, we present a new type of global adversarial attacks for assess…
RLVR maintains safety while improving reasoning capabilities in LLMs.
problem Safety-capability tradeoff in fine-tuning LLMs.
method Reinforcement Learning with Verifiable Rewards (RLVR) and theoretical analysis.
result RLVR can enhance reasoning while maintaining safety guardrails.
As machine learning systems become ubiquitous, there has been a surge of interest in interpretable machine learning: systems that provide explanation for their outputs. These explanations are often used to qualitatively assess other criteria such as safety or non-discrimination. However, despite the interest in interpr…
Topological parallax assesses AI models' geometric similarity to datasets for safety.
problem Ensuring AI models' robustness and safety in deep learning applications.
method Topological parallax compares a trained model to a reference dataset using Rips complexes and geodesic distortions.
result Topological parallax indicates whether a model shares similar multiscale geometric features with the dataset.
The increasing inclusion of Machine Learning (ML) models in safety critical systems like autonomous cars have led to the development of multiple model-based ML testing techniques. One common denominator of these testing techniques is their assumption that training programs are adequate and bug-free. These techniques on…
Autonomous vehicles rely on machine learning to solve challenging tasks in perception and motion planning. However, automotive software safety standards have not fully evolved to address the challenges of machine learning safety such as interpretability, verification, and performance limitations. In this paper, we revi…
Fine-tuning LLMs improves capability but harms safety, study finds.
problem Balancing capability and safety in LLM fine-tuning.
method Theoretical framework and numerical experiments for two safety-aware fine-tuning strategies.
result Characterization of fundamental limits of safety-capability trade-off in LLM fine-tuning.
RAGuard improves safety in LLMs for offshore wind maintenance.
problem Conventional LLMs fail with specialised or unexpected scenarios in offshore wind maintenance.
method Integrates safety-critical documents alongside technical manuals in RAG framework.
result RAGuard increases safety recall from almost 0% to over 50% while maintaining technical recall above 60%.
The paper explores new risk models for autonomous driving.
problem Risk management and actuarial modeling for autonomous vehicles.
method Examines technical difficulties and proposes a novel risk model.
result The new model better reflects real-world driving safety.
Survey of algorithms for testing AI-driven CPS safety.
problem Testing AI-driven CPS for safety in complex environments.
method Survey of applied algorithms for safety validation.
result Survey of existing tools and techniques for safety validation.
Predict real-time crash risks during hurricane evacuations using connected vehicle data.
problem Mitigate crash risks during hurricane evacuations by predicting high-risk locations.
method Used connected vehicle data to predict crash risks in real-time, considering weather and traffic features.
result Gaussian Process Boosting and Extreme Gradient Boosting models performed best, with recall of 0.91.
Safe imitation learning with a safety layer for flexible training.
problem Flexible yet safe imitation learning for complex tasks.
method Theory and modular method with a safety layer for continuous policy, adversarial training, and worst-case safety guarantees.
result Robustness advantage of safety layer during training compared to test time.
Deep learning methods are widely regarded as indispensable when it comes to designing perception pipelines for autonomous agents such as robots, drones or automated vehicles. The main reasons, however, for deep learning not being used for autonomous agents at large scale already are safety concerns. Deep learning appro…
Paper aims to ensure reliable detection of out-of-distribution data with certifiable worst-case guarantees.
problem Deep neural networks are overconfident with OOD inputs, posing safety risks.
method Enforces low confidence and bounds in an l∞-ball around OOD points using interval bound propagation (IBP). result Certifiable worst-case guarantees for OOD detection are possible without significant loss in accuracy.
TensorFI injects faults in TensorFlow programs to assess their reliability.
problem Ensuring reliability of machine learning systems in safety-critical domains.
method TensorFI is a flexible fault injection framework for TensorFlow applications.
result TensorFI evaluates the resilience of 12 ML programs, including autonomous vehicle DNNs.
Transportation agencies have an opportunity to leverage increasingly-available trajectory datasets to improve their analyses and decision-making processes. However, this data is typically purchased from vendors, which means agencies must understand its potential benefits beforehand in order to properly assess its value…
A new algorithm for identifying the best arm in linear feedback with safety constraints.
problem Identifying the best arm in linear feedback with safety constraints.
method A gap-based algorithm that ensures safety while minimizing sample complexity.
result The algorithm achieves meaningful sample complexity while ensuring safety.
Safety filter for unknown discrete-time systems with learned models and noise covariance.
problem Ensuring safety for unknown discrete-time linear systems with Gaussian noise.
method Develops a learning-based safety filter using empirical model and noise covariance, optimizing control actions to stay within safety constraints.
result Minimally modifies nominal control actions to ensure safety with high probability, tightening constraints as more data is collected.