搜索结果：Safety Research

Safety has become the central value around which dominant AI governance efforts are being shaped. Recently, this culminated in the publication of the International AI Safety Report, written by 96 experts of which 30 nominated by the Organisation for Economic Co-operation and Development (OECD), the European Union (EU), and the United Nations (UN). The report focuses on the safety risks of general-purpose AI and available technical mitigation approaches. In this response, informed by a system safety perspective, I refl ect on the key conclusions of the report, identifying fundamental issues in the currently dominant technical framing of AI safety and how this frustrates meaningful discourse and policy efforts to address safety comprehensively. The system safety discipline has dealt with the safety risks of software-based systems for many decades, and understands safety risks in AI systems as sociotechnical and requiring consideration of technical and non-technical factors and their interactions. The International AI Safety report does identify the need for system safety approaches. Lessons, concepts and methods from system safety indeed provide an important blueprint for overcoming

In Rust, unsafe code is the sole source of potential undefined behaviors. To avoid misuse, Rust developers should clarify the safety properties for each unsafe API. However, the community currently lacks a key standard for safety documentation: existing safety comments in the source code and safety documentation can be ad hoc and incomplete. This paper presents a tag-centric methodology for auditing the consistency and completeness of safety documentation. We first derive a taxonomy of Safety Tags to formalize natural-language requirements. Second, because API soundness frequently relies on struct invariants, we propose a set of empirical rules to systematically audit the structural consistency of safety documentation. We implemented this methodology in safety-tool, a static linter that automatically enforces structural consistency between local safety annotations and callee requirements. Our approach was applied to the Rust standard library, fixing documentation issues on 27 APIs with 61 safety tags and identifying safety tags that are applicable to 96.1% of the public unsafe APIs in libstd. Furthermore, we have formalized the tagging idea through a Rust RFC to the wider community

Despite the strong performance of large language models (LLMs) across diverse tasks, their susceptibility to adversarial attacks and unsafe content generation remains a significant obstacle to deployment, particularly in high-stakes settings. Addressing this challenge requires safety mechanisms that are both practically effective and theoretically grounded. In this paper, we introduce BarrierSteer, a novel inference-time framework that improves response safety by embedding learned nonlinear safety constraints directly into the model's latent representation space. BarrierSteer treats hidden-state safety classifiers as Control Barrier Functions (CBFs), enabling constraint-guided steering of unsafe latent trajectories during generation. By composing multiple safety constraints through efficient constraint merging without modifying the underlying LLM parameters, BarrierSteer preserves model utility. We provide theoretical results showing that applying CBFs in the latent space yields a principled, modular, and computationally efficient approach for steering with respect to learned safety constraints, with guarantees conditional on the learned barriers capturing the intended safety prope

Safety lies at the core of developing and deploying large language models (LLMs). However, previous safety benchmarks only concern the safety in one language, e.g. the majority language in the pretraining data such as English. In this work, we build the first multilingual safety benchmark for LLMs, XSafety, in response to the global deployment of LLMs in practice. XSafety covers 14 kinds of commonly used safety issues across 10 languages that span several language families. We utilize XSafety to empirically study the multilingual safety for 4 widely-used LLMs, including both close-API and open-source models. Experimental results show that all LLMs produce significantly more unsafe responses for non-English queries than English ones, indicating the necessity of developing safety alignment for non-English languages. In addition, we propose several simple and effective prompting methods to improve the multilingual safety of ChatGPT by evoking safety knowledge and improving cross-lingual generalization of safety alignment. Our prompting method can significantly reduce the ratio of unsafe responses from 19.1% to 9.7% for non-English queries. We release our data at https://github.com/Jar

Recent progress in generative modeling has made safety control a central challenge, yet existing approaches remain largely model-specific, requiring retraining or tailored interventions for each new architecture. In this work, we ask whether safety can be represented as a portable latent direction, learned once and reused across heterogeneous generators. We introduce the first framework for cross-model safety steering, in which a safety direction is estimated in a source LLM from paired safe-unsafe prompts, transported to a target generator through a lightweight alignment fitted on benign data alone, and applied at inference time. Crucially, our pipeline never accesses unsafe data on the target side, isolating whether safety can be transferred through shared representation geometry. Beyond a single global direction, we also identify a multi-vector extension that captures category-specific safety behaviors, enabling more selective control. We evaluate our approach in text-to-image and text-to-video generation across diverse source-target model pairs. Across models, transferred safety directions achieve ASR reduction and CLIP-Score/FID trade-offs comparable to directions learned nati

This paper develops a Geometric Analysis-based Route Safety Assessment (GARSA) framework to enhance the safety of marine vehicles navigating in restricted waters. Utilizing line and point geometric elements to define waterway boundaries, the framework enables the construction of a dynamic width characterization function to quantify spatial safety within complex restricted navigation spaces. An iterative method is developed to calculate the dynamic width characterization function, enabling an abstracted spatial property representation of waterways. Based on this, a navigational safety index that accounts for the overall route safety as well as the spatial constraints of local segments is created, enabling the selection of the safest waterway to pass through. Case studies in Hamburg Port and hypothetical waterways show that GARSA identifies spatial safety differences among routes. For example, GARSA assigned safety assessment values of 0.6 and 135 to two intra-port transit routes of 7,258 m and 7,845 m, respectively, with the higher value indicating a safer route. Overall, GARSA provides a quantitative basis for safety-oriented route decision-making of marine vehicles in restricted w

Multimodal large language models (MLLMs) are increasingly deployed as assistants that interact through text and images, making it crucial to evaluate contextual safety when risk depends on both the visual scene and the evolving dialogue. Existing contextual safety benchmarks are mostly single-turn and often miss how malicious intent can emerge gradually or how the same scene can support both benign and exploitative goals. We introduce the Multi-Turn Multimodal Contextual Safety Benchmark (MTMCS-Bench), a benchmark of realistic images and multi-turn conversations that evaluates contextual safety in MLLMs under two complementary settings, escalation-based risk and context-switch risk. MTMCS-Bench offers paired safe and unsafe dialogues with structured evaluation. It contains over 30 thousand multimodal (image+text) and unimodal (text-only) samples, with metrics that separately measure contextual intent recognition, safety-awareness on unsafe cases, and helpfulness on benign ones. Across eight open-source and seven proprietary MLLMs, we observe persistent trade-offs between contextual safety and utility, with models tending to either miss gradual risks or over-refuse benign dialogues.

The rapid rise of open-weight and open-source foundation models is intensifying the obligation and reshaping the opportunity to make AI systems safe. This paper reports outcomes from the Columbia Convening on AI Openness and Safety (San Francisco, 19 Nov 2024) and its six-week preparatory programme involving more than forty-five researchers, engineers, and policy leaders from academia, industry, civil society, and government. Using a participatory, solutions-oriented process, the working groups produced (i) a research agenda at the intersection of safety and open source AI; (ii) a mapping of existing and needed technical interventions and open source tools to safely and responsibly deploy open foundation models across the AI development workflow; and (iii) a mapping of the content safety filter ecosystem with a proposed roadmap for future research and development. We find that openness -- understood as transparent weights, interoperable tooling, and public governance -- can enhance safety by enabling independent scrutiny, decentralized mitigation, and culturally plural oversight. However, significant gaps persist: scarce multimodal and multilingual benchmarks, limited defenses agai

Large Language Models' safety-aligned behaviors, such as refusing harmful queries, can be represented by linear directions in activation space. Previous research modeled safety behavior with a single direction, limiting mechanistic understanding to an isolated safety feature. In this work, we discover that safety-aligned behavior is jointly controlled by multi-dimensional directions. Namely, we study the vector space of representation shifts during safety fine-tuning on Llama 3 8B for refusing jailbreaks. By studying orthogonal directions in the space, we first find that a dominant direction governs the model's refusal behavior, while multiple smaller directions represent distinct and interpretable features like hypothetical narrative and role-playing. We then measure how different directions promote or suppress the dominant direction, showing the important role of secondary directions in shaping the model's refusal representation. Finally, we demonstrate that removing certain trigger tokens in harmful queries can mitigate these directions to bypass the learned safety capability, providing new insights on understanding safety alignment vulnerability from a multi-dimensional perspec

A system safety case is a compelling, comprehensible, and valid argument about the satisfaction of the safety goals of a given system operating in a given environment supported by convincing evidence. Since the publication of UL 4600 in 2020, safety cases have become a best practice for measuring, managing, and communicating the safety of autonomous vehicles (AVs). Although UL 4600 provides guidance on how to build the safety case for an AV, the complexity of AVs and their operating environments, the novelty of the used technology, the need for complying with various regulations and technical standards, and for addressing cybersecurity concerns and ethical considerations make the development of safety cases for AVs challenging. To this end, safety case frameworks have been proposed that bring strategies, argument templates, and other guidance together to support the development of a safety case. This paper introduces the Open Autonomy Safety Case Framework, developed over years of work with the autonomous vehicle industry, as a roadmap for how AVs can be deployed safely and responsibly.

Aviation safety is paramount, demanding precise analysis of safety occurrences during different flight phases. This study employs Natural Language Processing (NLP) and Deep Learning models, including LSTM, CNN, Bidirectional LSTM (BLSTM), and simple Recurrent Neural Networks (sRNN), to classify flight phases in safety reports from the Australian Transport Safety Bureau (ATSB). The models exhibited high accuracy, precision, recall, and F1 scores, with LSTM achieving the highest performance of 87%, 88%, 87%, and 88%, respectively. This performance highlights their effectiveness in automating safety occurrence analysis. The integration of NLP and Deep Learning technologies promises transformative enhancements in aviation safety analysis, enabling targeted safety measures and streamlined report handling.

We propose a new framework to facilitate dynamic assurance within a safety case approach by associating safety performance measurement with the core assurance artifacts of a safety case. The focus is mainly on the safety architecture, whose underlying risk assessment model gives the concrete link from safety measurement to operational risk. Using an aviation domain example of autonomous taxiing, we describe our approach to derive safety indicators and revise the risk assessment based on safety measurement. We then outline a notion of consistency between a collection of safety indicators and the safety case, as a formal basis for implementing the proposed framework in our tool, AdvoCATE.

Autonomous vehicles (AVs) have significantly advanced in real-world deployment in recent years, yet safety continues to be a critical barrier to widespread adoption. Traditional functional safety approaches, which primarily verify the reliability, robustness, and adequacy of AV hardware and software systems from a vehicle-centric perspective, do not sufficiently address the AV's broader interactions and behavioral impact on the surrounding traffic environment. To overcome this limitation, we propose a paradigm shift toward behavioral safety, a comprehensive approach focused on evaluating AV responses and interactions within traffic environment. To systematically assess behavioral safety, we introduce a third-party AV safety assessment framework comprising two complementary evaluation components: Driver Licensing Test and Driving Intelligence Test. The Driver Licensing Test evaluates AV's reactive behaviors under controlled scenarios, ensuring basic behavioral competency. In contrast, the Driving Intelligence Test assesses AV's interactive behaviors within naturalistic traffic conditions, quantifying the frequency of safety-critical events to deliver statistically meaningful safety

We develop a data-driven approach for runtime safety monitoring in flight testing, where pilots perform maneuvers on aircraft with uncertain parameters. Because safety violations can arise unexpectedly as a result of these uncertainties, pilots need clear, preemptive criteria to abort the maneuver in advance of safety violation. To solve this problem, we use offline stochastic trajectory simulation to learn a calibrated statistical model of the short-term safety risk facing pilots. We use flight testing as a motivating example for data-driven learning/monitoring of safety due to its inherent safety risk, uncertainty, and human-interaction. However, our approach consists of three broadly-applicable components: a model to predict future state from recent observations, a nearest neighbor model to classify the safety of the predicted state, and classifier calibration via conformal prediction. We evaluate our method on a flight dynamics model with uncertain parameters, demonstrating its ability to reliably identify unsafe scenarios, match theoretical guarantees, and outperform baseline approaches in preemptive classification of risk.

This paper presents a scientometric analysis of research output from the University of Lagos, focusing on the two decades spanning 2004 to 2023. Using bibliometric data retrieved from the Web of Science, we examine trends in publication volume, collaboration patterns, citation impact, and the most prolific authors, departments, and research domains at the university. The study reveals a consistent increase in research productivity, with the highest publication output recorded in 2023. Health Sciences, Engineering, and Social Sciences are identified as dominant fields, reflecting the university's interdisciplinary research strengths. Collaborative efforts, both locally and internationally, show a positive correlation with higher citation impact, with the United States and the United Kingdom being the leading international collaborators. Notably, open-access publications account for a significant portion of the university's research output, enhancing visibility and citation rates. The findings offer valuable insights into the university's research performance over the past two decades, providing a foundation for strategic planning and policy formulation to foster research excellence

Ensuring the safe alignment of large language models (LLMs) with human values is critical as they become integral to applications like translation and question answering. Current alignment methods struggle with dynamic user intentions and complex objectives, making models vulnerable to generating harmful content. We propose Safety Arithmetic, a training-free framework enhancing LLM safety across different scenarios: Base models, Supervised fine-tuned models (SFT), and Edited models. Safety Arithmetic involves Harm Direction Removal to avoid harmful content and Safety Alignment to promote safe responses. Additionally, we present NoIntentEdit, a dataset highlighting edit instances that could compromise model safety if used unintentionally. Our experiments show that Safety Arithmetic significantly improves safety measures, reduces over-safety, and maintains model utility, outperforming existing methods in ensuring safe content generation.

Modern cyber-physical systems are operated by complex software that increasingly takes over safety-critical functions. Software enables rapid iterations and continuous delivery of new functionality that meets the ever-changing expectations of users. As high-speed development requires discipline, rigor, and automation, software factories are used. These entail methods and tools used for software development, such as build systems and pipelines. To keep up with the rapid evolution of software, we need to bridge the disconnect in methods and tools between software development and safety engineering today. We need to invest more in formality upfront - capturing safety work products in semantically rich models that are machine-processable, defining automatic consistency checks, and automating the generation of documentation - to benefit later. Transferring best practices from software to safety engineering is worth exploring. We advocate for safety factories, which integrate safety tooling and methods into software development pipelines.

The development of safety-oriented research and applications requires fine-grain vehicle trajectories that not only have high accuracy, but also capture substantial safety-critical events. However, it would be challenging to satisfy both these requirements using the available vehicle trajectory datasets do not have the capacity to satisfy both.This paper introduces the CitySim dataset that has the core objective of facilitating safety-oriented research and applications. CitySim has vehicle trajectories extracted from 1140 minutes of drone videos recorded at 12 locations. It covers a variety of road geometries including freeway basic segments, signalized intersections, stop-controlled intersections, and control-free intersections. CitySim was generated through a five-step procedure that ensured trajectory accuracy. The five-step procedure included video stabilization, object filtering, multi-video stitching, object detection and tracking, and enhanced error filtering. Furthermore, CitySim provides the rotated bounding box information of a vehicle, which was demonstrated to improve safety evaluations. Compared with other video-based critical events, including cut-in, merge, and diver

Demographic data collection is essential in education research, as demographic data allows researchers to better describe the participant population they study and to contextualize findings. However, current research practices for neurodiversity demographics often rely on prescriptive methods (e.g., requiring participants to report official diagnoses) rather than allowing participants to self-identify. This approach can: a) not allow participants to express their intersecting identities in ways that are authentic; and b) limit trustworthiness and reliability of the data and interpretation. In addition, inconsistent dissemination and representation of demographic data across studies hinder the accessibility and usability of this work. Through a literature review of neurodivergent student experiences with learning and performing STEM, we identified widespread discrepancies in how demographic information is collected and reported. This paper explores how neurodivergent identities can be more accurately and inclusively represented in education research. We present findings of a thematic analysis on the ways neurodivergent demographic data collection is done in the literature using data