Large Language Models (LLMs) have raised increasing concerns about their misuse in generating hate speech. Among all the efforts to address this issue, hate speech detectors play a crucial role. However, the effectiveness of different detectors against LLM-generated hate speech remains largely unknown. In this paper, we propose HateBench, a framework for benchmarking hate speech detectors on LLM-generated hate speech. We first construct a hate speech dataset of 7,838 samples generated by six widely-used LLMs covering 34 identity groups, with meticulous annotations by three labelers. We then assess the effectiveness of eight representative hate speech detectors on the LLM-generated dataset. Our results show that while detectors are generally effective in identifying LLM-generated hate speech, their performance degrades with newer versions of LLMs. We also reveal the potential of LLM-driven hate campaigns, a new threat that LLMs bring to the field of hate speech detection. By leveraging advanced techniques like adversarial attacks and model stealing attacks, the adversary can intentionally evade the detector and automate hate campaigns online. The most potent adversarial attack achieves an attack success rate of 0.966, and its attack efficiency can be further improved by $13-21\times$ through model stealing attacks with acceptable attack performance. We hope our study can serve as a call to action for the research community and platform moderators to fortify defenses against these emerging threats.

Abstract--In the near future, autonomous space systems will compose a large number of the spacecraft being deployed. Their tasks will involve autonomous rendezvous and proximity operations with large structures, such as inspections or assembly of orbiting space stations and maintenance and human-assistance tasks over shared workspaces. To promote replicable and reliable scientific results for autonomous control of spacecraft, we present the design of a space systems laboratory based on open-source and modular software and hardware. The simulation software provides a software-in-the-loop (SITL) architecture that seamlessly transfers simulated results to the ATMOS platforms, developed for testing of multi-agent autonomy schemes for microgravity. The manuscript presents the KTH space systems laboratory facilities and the ATMOS platform as open-source hardware and software contributions. To the left, we see the tethers of the low-pressure compressor system. Software and hardware contributions can be found in: 1. PX4Space: Athens [6] proposed a similar test bed, where the platforms https://atmos.discower.io The facility also provides a vision-based I. This The space sector has experienced significant growth [1] in facility was recently upgraded to more modern avionics, motion the last decade, in part due to the decreased costs of access to capture ground-truth positioning, and robotics communication space through multiple commercial operators [2], but also due software through the Robotics Operating System (ROS) [7]. to the maturation of existing technologies and, consequently, Stanford University's Autonomous Systems Laboratory freeflyer reduced pricing of the deployed equipment. In the last twenty testbed [8], [9], [10] uses a similar, round platform as to thirty years, a few academic and industrial research facilities a free-flying robotic system for path planning, docking and have been created to test space systems by replicating motion capturing of space systems, paired with an open-source Python in microgravity on Earth.

Resource-constrained edge deployments demand AI solutions that balance high performance with stringent compute, memory, and energy limitations. In this survey, we present a comprehensive overview of the primary strategies for accelerating deep learning models under such constraints. First, we examine model compression techniques-pruning, quantization, tensor decomposition, and knowledge distillation-that streamline large models into smaller, faster, and more efficient variants. Next, we explore Neural Architecture Search (NAS), a class of automated methods that discover architectures inherently optimized for particular tasks and hardware budgets. We then discuss compiler and deployment frameworks, such as TVM, TensorRT, and OpenVINO, which provide hardware-tailored optimizations at inference time. By integrating these three pillars into unified pipelines, practitioners can achieve multi-objective goals, including latency reduction, memory savings, and energy efficiency-all while maintaining competitive accuracy. We also highlight emerging frontiers in hierarchical NAS, neurosymbolic approaches, and advanced distillation tailored to large language models, underscoring open challenges like pre-training pruning for massive networks. Our survey offers practical insights, identifies current research gaps, and outlines promising directions for building scalable, platform-independent frameworks to accelerate deep learning models at the edge.

Prior research on training grounded factuality classification models to detect hallucinations in large language models (LLMs) has relied on public natural language inference (NLI) data and synthetic data. However, conventional NLI datasets are not well-suited for document-level reasoning, which is critical for detecting LLM hallucinations. Recent approaches to document-level synthetic data generation involve iteratively removing sentences from documents and annotating factuality using LLM-based prompts. While effective, this method is computationally expensive for long documents and limited by the LLM's capabilities. In this work, we analyze the differences between existing synthetic training data used in state-of-the-art models and real LLM output claims. Based on our findings, we propose a novel approach for synthetic data generation, CG2C, that leverages multi-hop reasoning on context graphs extracted from documents. Our fact checker model, FactCG, demonstrates improved performance with more connected reasoning, using the same backbone models. Experiments show it even outperforms GPT-4-o on the LLM-Aggrefact benchmark with much smaller model size.

Large Language Models (LLMs) have recently gained attention due to their ability to understand and generate sophisticated human-like content. However, ensuring their safety is paramount as they might provide harmful and unsafe responses. Existing LLM testing frameworks address various safety-related concerns (e.g., drugs, terrorism, animal abuse) but often face challenges due to unbalanced and obsolete datasets. In this paper, we present ASTRAL, a tool that automates the generation and execution of test cases (i.e., prompts) for testing the safety of LLMs. First, we introduce a novel black-box coverage criterion to generate balanced and diverse unsafe test inputs across a diverse set of safety categories as well as linguistic writing characteristics (i.e., different style and persuasive writing techniques). Second, we propose an LLM-based approach that leverages Retrieval Augmented Generation (RAG), few-shot prompting strategies and web browsing to generate up-to-date test inputs. Lastly, similar to current LLM test automation techniques, we leverage LLMs as test oracles to distinguish between safe and unsafe test outputs, allowing a fully automated testing approach. We conduct an extensive evaluation on well-known LLMs, revealing the following key findings: i) GPT3.5 outperforms other LLMs when acting as the test oracle, accurately detecting unsafe responses, and even surpassing more recent LLMs (e.g., GPT-4), as well as LLMs that are specifically tailored to detect unsafe LLM outputs (e.g., LlamaGuard); ii) the results confirm that our approach can uncover nearly twice as many unsafe LLM behaviors with the same number of test inputs compared to currently used static datasets; and iii) our black-box coverage criterion combined with web browsing can effectively guide the LLM on generating up-to-date unsafe test inputs, significantly increasing the number of unsafe LLM behaviors.

The rapid deployment of Artificial Intelligence (AI) in critical digital infrastructure introduces significant risks, necessitating a robust framework for systematically collecting AI incident data to prevent future incidents. Existing databases lack the granularity as well as the standardized structure required for consistent data collection and analysis, impeding effective incident management. This work proposes a standardized schema and taxonomy for AI incident databases, addressing these challenges by enabling detailed and structured documentation of AI incidents across sectors. Key contributions include developing a unified schema, introducing new fields such as incident severity, causes, and harms caused, and proposing a taxonomy for classifying AI incidents in critical digital infrastructure. The proposed solution facilitates more effective incident data collection and analysis, thus supporting evidence-based policymaking, enhancing industry safety measures, and promoting transparency. This work lays the foundation for a coordinated global response to AI incidents, ensuring trust, safety, and accountability in using AI across regions.

Modern machine learning produces models that are impossible for users or developers to fully understand -- raising concerns about trust, oversight and human dignity. Transparency and explainability methods aim to provide some help in understanding models, but it remains challenging for developers to design explanations that are understandable to target users and effective for their purpose. Emerging guidelines and regulations set goals but may not provide effective actionable guidance to developers. In a controlled experiment with 124 participants, we investigate whether and how specific forms of policy guidance help developers design explanations for an ML-powered screening tool for diabetic retinopathy. Contrary to our expectations, we found that participants across the board struggled to produce quality explanations, comply with the provided policy requirements for explainability, and provide evidence of compliance. We posit that participant noncompliance is in part due to a failure to imagine and anticipate the needs of their audience, particularly non-technical stakeholders. Drawing on cognitive process theory and the sociological imagination to contextualize participants' failure, we recommend educational interventions.

The proliferation of powerful AI capabilities and systems necessitates a commensurate focus on user trust. We introduce the Trust Calibration Maturity Model (TCMM) to capture and communicate the maturity of AI system trustworthiness. The TCMM scores maturity along 5 dimensions that drive user trust: Performance Characterization, Bias & Robustness Quantification, Transparency, Safety & Security, and Usability. Information captured in the TCMM can be presented along with system performance information to help a user to appropriately calibrate trust, to compare requirements with current states of development, and to clarify trustworthiness needs. We present the TCMM and demonstrate its use on two AI system-target task pairs.

Adversarial robustness in quantum classifiers is a critical area of study, providing insights into their performance compared to classical models and uncovering potential advantages inherent to quantum machine learning. In the NISQ era of quantum computing, circuit cutting is a notable technique for simulating circuits that exceed the qubit limitations of current devices, enabling the distribution of a quantum circuit's execution across multiple quantum processing units through classical communication. We examine how partitioning quantum classifiers through circuit cutting increase their susceptibility to adversarial attacks, establishing a link between attacking the state preparation channels in wire cutting and implementing adversarial gates within intermediate layers of a quantum classifier. We then proceed to study the latter problem from both a theoretical and experimental perspective.

Reinforcement Learning from Human Feedback (RLHF) is commonly employed to tailor models to human preferences, especially to improve the safety of outputs from large language models (LLMs). Traditionally, this method depends on selecting preferred responses from pairs. However, due to the variability in human opinions and the challenges in directly comparing two responses, there is an increasing trend towards fine-grained annotation approaches that evaluate responses using multiple targeted metrics or rules. The challenge lies in efficiently choosing and applying these rules to handle the diverse range of preference data. In this paper, we propose a dynamic method that adaptively selects the most important rules for each response pair. We introduce a mathematical framework that utilizes the maximum discrepancy across paired responses and demonstrate theoretically that this approach maximizes the mutual information between the rule-based annotations and the underlying true preferences. We then train an 8B reward model using this adaptively labeled preference dataset and assess its efficacy using RewardBench. As of January 25, 2025, our model achieved the highest safety performance on the leaderboard, surpassing various larger models.