Search: “safety”

arXiv:2606.28332v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used for medical and health-related questions, yet their safety in high-risk medical scenarios remains p

arXiv:2606.25750v1 Announce Type: cross Abstract: Safety evaluation of large language models (LLMs) is commonly performed by querying models with unsafe or jailbreak prompts and judging whether their

arXiv:2606.24245v1 Announce Type: cross Abstract: Large language model (LLM) agents increasingly automate complex tasks by integrating language models with external tools and environments. However, th

arXiv:2606.10154v1 Announce Type: new Abstract: Quantized checkpoints are often screened first with quality metrics and only later, if at all, with direct safety tests. This paper audits that shortcut

arXiv:2606.08281v1 Announce Type: new Abstract: Physical human-robot interaction (pHRI) demands simultaneous trajectory accuracy and compliant safety under unplanned contact. Classical impedance contr

arXiv:2606.08414v1 Announce Type: cross Abstract: Diffusion policies have achieved remarkable success in robotic manipulation, yet they often fail to satisfy strict physical constraints required for s

arXiv:2606.05461v1 Announce Type: new Abstract: Safety standards for ML-based autonomous driving specify the kind of evidence an assurance case must contain (directed cause-and-effect chains, quantifi

arXiv:2606.05177v1 Announce Type: new Abstract: Existing multimodal safety benchmarks focus solely on visual inputs and cannot assess Omni Large Language Models (LLMs) that process vision, audio, and

arXiv:2605.30368v1 Announce Type: cross Abstract: Surrogate Safety Measures (SSMs) are extensively utilised in the evaluation of traffic risk in automated driving contexts. However, the majority of SS

arXiv:2605.29659v1 Announce Type: cross Abstract: Real-time safety filtering for large language model (LLM) applications requires classifiers that can detect unsafe prompts, toxic language, jailbreak

arXiv:2605.29708v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) LLMs rely on sparse, router-driven expert activation, yet how safety alignment interacts with routed expert specialization rema

arXiv:2605.28664v1 Announce Type: cross Abstract: Safety detection models require examples of HHH (Helpful, Harmless, Honest)-violating outputs for robust generalization, however such examples are sca

arXiv:2601.01627v3 Announce Type: replace-cross Abstract: As Large Language Models (LLMs) are increasingly deployed in healthcare field, it becomes essential to carefully evaluate their medical safety

arXiv:2605.28367v1 Announce Type: new Abstract: Safe physical interaction is critical for deploying robotic manipulators in human-robot interaction and contact-rich tasks, where uncertainty, external

arXiv:2603.07211v2 Announce Type: replace Abstract: Direct Preference Optimization (DPO) has become a standard framework for safety alignment, but its reliance on pairwise preference updates makes tra

arXiv:2601.21845v2 Announce Type: replace Abstract: Meta reinforcement learning (RL) allows agents to leverage experience across a distribution of tasks on which the agent can train at will, enabling

arXiv:2605.27155v1 Announce Type: cross Abstract: Testing object detectors in safety-critical domains requires semantically meaningful probes beyond pixel-level corruptions. We present SemProbe, a too

arXiv:2512.11551v2 Announce Type: replace Abstract: The growing number of road users has significantly increased the risk of accidents in recent years. Vulnerable Road Users (VRUs) are particularly at

arXiv:2605.21712v1 Announce Type: new Abstract: Transportation safety analysis requires integrating crash records, roadway attributes, and geospatial data through GIS-based workflows, but access remai

arXiv:2605.21109v1 Announce Type: new Abstract: Reliable confidence estimates are important for safely deploying vision-based controllers in autonomous racing, where safety predictions must be derived

arXiv:2605.20241v1 Announce Type: cross Abstract: Prompt-level safety probes for large language models use hidden-state representations to separate safe from unsafe prompts, but strong average detecti

arXiv:2605.20625v1 Announce Type: cross Abstract: Advanced Air Mobility (AAM) operations are expected to significantly increase aerial traffic in urban airspace, requiring autonomous traffic managemen

arXiv:2605.17413v1 Announce Type: cross Abstract: Safety-aligned language models often refuse cybersecurity requests whose wording resembles misuse, even when the task is authorized and defensive. Thi

arXiv:2503.02574v2 Announce Type: replace-cross Abstract: In this paper, we argue that current safety alignment research efforts for large language models are hindered by many intertwined sources of n

arXiv:2603.18178v2 Announce Type: replace-cross Abstract: The rapid growth of ego-centric dashcam footage presents a major challenge for detecting safety-critical events such as collisions and near-co

arXiv:2603.17915v2 Announce Type: replace-cross Abstract: As large language models (LLMs) are deployed in multilingual settings, their safety behavior in culturally diverse, low-resource languages rem

arXiv:2509.26100v2 Announce Type: replace Abstract: The rapid integration of Large Language Models (LLMs) into high-stakes domains necessitates reliable safety and compliance evaluation. However, exis

arXiv:2605.15164v1 Announce Type: cross Abstract: This position paper argues that behavioural assurance, even when carefully designed, is being asked to carry safety claims it cannot verify. AI govern

arXiv:2605.14152v1 Announce Type: cross Abstract: Safety evaluations for large language models (LLMs) increasingly target high-stakes National Security and Public Safety (NSPS) risks, yet multilingual

arXiv:2605.12726v1 Announce Type: new Abstract: Final-token safety probes monitor a single hidden state after prompt prefill, but jailbreak prompts can contain probe-visible unsafe evidence distribute

arXiv:2602.07668v2 Announce Type: replace Abstract: The looking-in-looking-out (LILO) framework has enabled intelligent vehicle applications that understand both the outside scene and the driver state

arXiv:2605.11769v1 Announce Type: new Abstract: Air Traffic Control (ATC) is a safety-critical domain in which incorrect interpretation of instructions may lead to severe operational consequences. Whi

arXiv:2605.09045v1 Announce Type: new Abstract: Agentic frameworks are the software layer through which AI agents act in the world. Existing safety methods intervene on the model and therefore remain

arXiv:2605.02900v1 Announce Type: cross Abstract: Embodied Artificial Intelligence (Embodied AI) integrates perception, cognition, planning, and interaction into agents that operate in open-world, saf

arXiv:2509.00673v2 Announce Type: replace Abstract: We investigate the efficacy of Large Language Models (LLMs) in detecting implicit and explicit hate speech, examining how models with minimal safety

arXiv:2605.01761v1 Announce Type: new Abstract: Text-to-Video (T2V) models have demonstrated remarkable capability in generating temporally coherent videos from natural language prompts, yet they also

arXiv:2505.17342v2 Announce Type: replace Abstract: Safe Reinforcement Learning (SafeRL) is the subfield of reinforcement learning that explicitly deals with safety constraints during the learning and

arXiv:2604.26052v1 Announce Type: new Abstract: Safety evaluations of large language models (LLMs) typically report binary outcomes such as attack success rate, refusal rate, or harmful/not-harmful re

arXiv:2601.18491v2 Announce Type: replace Abstract: The rise of AI agents introduces complex safety and security challenges arising from autonomous tool use and environmental interactions. Current gua

arXiv:2604.18946v1 Announce Type: new Abstract: Large reasoning models (LRMs) achieve strong performance on complex reasoning tasks but often generate harmful responses to malicious user queries. This

arXiv:2604.18627v1 Announce Type: new Abstract: Ensuring human safety is of paramount importance in warehouse environments that feature mixed traffic of human workers and autonomous mobile robots (AMR

arXiv:2604.19001v1 Announce Type: new Abstract: Large reasoning models (LRMs) produce complex, multi-step reasoning traces, yet safety evaluation remains focused on final outputs, overlooking how harm

arXiv:2604.16424v1 Announce Type: cross Abstract: State-Space Models (SSMs) -- structured SSMs (S4, S4D, DSS, S5), selective SSMs (Mamba, Mamba-2), and hybrid architectures (Jamba) -- are deployed in

arXiv:2601.04043v2 Announce Type: replace Abstract: As Multimodal Large Language Models (MLLMs) become an indispensable assistant in human life, the unsafe content generated by MLLMs poses a danger to

arXiv:2509.18127v3 Announce Type: replace-cross Abstract: Sparse autoencoders (SAEs) enable interpretability research by decomposing entangled model activations into monosemantic features. However, un

arXiv:2604.09189v1 Announce Type: cross Abstract: LLMs internalize safety policies through RLHF, yet these policies are never formally specified and remain difficult to inspect. Existing benchmarks ev