Abstract--Large Language Models are fundamental actors in the modern IT landscape dominated by AI solutions. However, security threats associated with them might prevent their reliable adoption in critical application scenarios such as government organizations and medical institutions. For this reason, commercial LLMs typically undergo a sophisticated censoring mechanism to eliminate any harmful output they could possibly produce. These mechanisms maintain the integrity of LLM alignment by guaranteeing that the models respond safely and ethically. In response to this, attacks on LLMs are a significant threat to such protections, and many previous approaches have already demonstrated their effectiveness across diverse domains. Existing LLM attacks mostly adopt a generate-and-test strategy to craft malicious input. T o improve the comprehension of censoring mechanisms and design a targeted attack, we propose an Explainable-AI solution that comparatively analyzes the behavior of censored and uncensored models to derive unique exploitable alignment patterns. Then, we propose XBreaking, a novel approach that exploits these unique patterns to break the security and alignment constraints of LLMs by targeted noise injection. Our thorough experimental campaign returns important insights about the censoring mechanisms and demonstrates the effectiveness and performance of our approach. Nowadays, Large Language Models (LLMs, for short) represent the most promising and relevant advancement in the field of Artificial Intelligence. These complex deep learning models are trained on massive datasets that cover almost all aspects of people's daily lives, thus granting them the capability of generating, understanding, and processing human language. For this reason, their integration as support tools is becoming pervasive with applications spanning from text editor and proofreading to virtual assistant and personalized text generation. However, the diffusion of this technology, especially in critical domains such as government organizations and medical institutions, imposes the assessment of their security and privacy characteristics.

Providing high-quality feedback to Large Language Models (LLMs) on a diverse training distribution can be difficult and expensive, and providing feedback only on a narrow distribution can result in unintended generalizations. To better leverage narrow training data, we propose contrastive weight steering, a simple post-training method that edits the model parameters using weight arithmetic. We isolate a behavior direction in weight-space by subtracting the weight deltas from two small fine-tunes -- one that induces the desired behavior and another that induces its opposite -- and then add or remove this direction to modify the model's weights. We apply this technique to mitigate sycophancy and induce misalignment, and find that weight steering often generalizes further than activation steering, achieving stronger out-of-distribution behavioral control before degrading general capabilities. We also show that, in the context of task-specific fine-tuning, weight steering can partially mitigate undesired behavioral drift: it can reduce sycophancy and under-refusals introduced during fine-tuning while preserving task performance gains. Finally, we provide preliminary evidence that emergent misalignment can be detected by measuring the similarity between fine-tuning updates and an "evil" weight direction, suggesting that it may be possible to monitor the evolution of weights during training and detect rare misaligned behaviors that never manifest during training or evaluations.

AI has proven highly successful at urban planning analysis -- learning patterns from data to predict future conditions. The next frontier is AI-assisted decision-making: agents that recommend sites, allocate resources, and evaluate trade-offs while reasoning transparently about constraints and stakeholder values. Recent breakthroughs in reasoning AI -- CoT prompting, ReAct, and multi-agent collaboration frameworks -- now make this vision achievable. This position paper presents the Agentic Urban Planning AI Framework for reasoning-capable planning agents that integrates three cognitive layers (Perception, Foundation, Reasoning) with six logic components (Analysis, Generation, Verification, Evaluation, Collaboration, Decision) through a multi-agents collaboration framework. We demonstrate why planning decisions require explicit reasoning capabilities that are value-based (applying normative principles), rule-grounded (guaranteeing constraint satisfaction), and explainable (generating transparent justifications) -- requirements that statistical learning alone cannot fulfill. We compare reasoning agents with statistical learning, present a comprehensive architecture with benchmark evaluation metrics, and outline critical research challenges. This framework shows how AI agents can augment human planners by systematically exploring solution spaces, verifying regulatory compliance, and deliberating over trade-offs transparently -- not replacing human judgment but amplifying it with computational reasoning capabilities.

Criminal justice administrative data contain only a limited amount of information about the committed offense. However, there is an unused source of extensive information in continental European courts' decisions: descriptions of criminal behaviors in verdicts by which offenders are found guilty. In this paper, we study the feasibility of extracting these descriptions from publicly available court decisions from Slovakia. We use two different approaches for retrieval: regular expressions and large language models (LLMs). Our baseline was a simple method employing regular expressions to identify typical words occurring before and after the description. The advanced regular expression approach further focused on "sparing" and its normalization (insertion of spaces between individual letters), typical for delineating the description. The LLM approach involved prompting the Gemini Flash 2.0 model to extract the descriptions using predefined instructions. Although the baseline identified descriptions in only 40.5% of verdicts, both methods significantly outperformed it, achieving 97% with advanced regular expressions and 98.75% with LLMs, and 99.5% when combined. Evaluation by law students showed that both advanced methods matched human annotations in about 90% of cases, compared to just 34.5% for the baseline. LLMs fully matched human-labeled descriptions in 91.75% of instances, and a combination of advanced regular expressions with LLMs reached 92%.

Large language models (LLMs) are typically aligned to a universal set of safety and usage principles intended for broad public acceptability. Yet, real-world applications of LLMs often take place within organizational ecosystems shaped by distinctive corporate policies, regulatory requirements, use cases, brand guidelines, and ethical commitments. This reality highlights the need for rigorous and comprehensive evaluation of LLMs with pluralistic alignment goals, an alignment paradigm that emphasizes adaptability to diverse user values and needs. In this work, we present PLURALISTIC BEHAVIOR SUITE (PBSUITE), a dynamic evaluation suite designed to systematically assess LLMs' capacity to adhere to pluralistic alignment specifications in multi-turn, interactive conversations. PBSUITE consists of (1) a diverse dataset of 300 realistic LLM behavioral policies, grounded in 30 industries; and (2) a dynamic evaluation framework for stress-testing model compliance with custom behavioral specifications under adversarial conditions. Using PBSUITE, We find that leading open- and closed-source LLMs maintain robust adherence to behavioral policies in single-turn settings (less than 4% failure rates), but their compliance weakens substantially in multi-turn adversarial interactions (up to 84% failure rates). These findings highlight that existing model alignment and safety moderation methods fall short in coherently enforcing pluralistic behavioral policies in real-world LLM interactions. Our work contributes both the dataset and analytical framework to support future research toward robust and context-aware pluralistic alignment techniques.

Large Language Models (LLMs) distinguish themselves by quickly delivering information and providing personalized responses through natural language prompts. However, they also infer user demographics, which can raise ethical concerns about bias and implicit personalization and create an echo chamber effect. This study aims to explore how inferred political views impact the responses of ChatGPT globally, regardless of the chat session. We also investigate how custom instruction and memory features alter responses in ChatGPT, considering the influence of political orientation. We developed three personas (two politically oriented and one neutral), each with four statements reflecting their viewpoints on DEI programs, abortion, gun rights, and vaccination. We convey the personas' remarks to ChatGPT using memory and custom instructions, allowing it to infer their political perspectives without directly stating them. We then ask eight questions to reveal differences in worldview among the personas and conduct a qualitative analysis of the responses. Our findings indicate that responses are aligned with the inferred political views of the personas, showing varied reasoning and vocabulary, even when discussing similar topics. We also find the inference happening with explicit custom instructions and the implicit memory feature in similar ways. Analyzing response similarities reveals that the closest matches occur between the democratic persona with custom instruction and the neutral persona, supporting the observation that ChatGPT's outputs lean left.

We introduce POLIS-Bench, the first rigorous, systematic evaluation suite designed for LLMs operating in governmental bilingual policy scenarios. Compared to existing benchmarks, POLIS-Bench introduces three major advancements. (i) Up-to-date Bilingual Corpus: We construct an extensive, up-to-date policy corpus that significantly scales the effective assessment sample size, ensuring relevance to current governance practice. (ii) Scenario-Grounded Task Design: We distill three specialized, scenario-grounded tasks -- Clause Retrieval & Interpretation, Solution Generation, and the Compliance Judgmen--to comprehensively probe model understanding and application. (iii) Dual-Metric Evaluation Framework: We establish a novel dual-metric evaluation framework combining semantic similarity with accuracy rate to precisely measure both content alignment and task requirement adherence. A large-scale evaluation of over 10 state-of-the-art LLMs on POLIS-Bench reveals a clear performance hierarchy where reasoning models maintain superior cross-task stability and accuracy, highlighting the difficulty of compliance tasks. Furthermore, leveraging our benchmark, we successfully fine-tune a lightweight open-source model. The resulting POLIS series models achieves parity with, or surpasses, strong proprietary baselines on multiple policy subtasks at a significantly reduced cost, providing a cost-effective and compliant path for robust real-world governmental deployment.

This paper examines the role of public interest litigation in promoting accountability for AI and automated decision-making (ADM) in Australia. Since ADM regulation faces geopolitical headwinds, effective governance will have to rely at least in part on the enforcement of existing laws. Drawing on interviews with Australian public interest litigators, technology policy activists, and technology law scholars, the paper positions public interest litigation as part of a larger ecosystem for transparency, accountability and justice with respect to ADM. It builds on one participant's characterisation of litigation about ADM as an exercise in legal retrofitting: adapting old laws to new circumstances. The paper's primary contribution is to aggregate, organise and present original insights on pragmatic strategies and tactics for effective public interest litigation about ADM. Naturally, it also contends with the limits of these strategies, and of the Australian legal system. Where limits are, however, capable of being overcome, the paper presents findings on urgent needs: the enabling institutional arrangements without which effective litigation and accountability will falter. The paper is relevant to law and technology scholars; individuals and groups harmed by ADM; public interest litigators and technology lawyers; civil society and advocacy organisations; and policymakers.

Pre-trained Time Series Foundational Models (TSFMs) represent a significant advance, capable of forecasting diverse time series with complex characteristics, including varied seasonalities, trends, and long-range dependencies. Despite their primary goal of universal time series forecasting, their efficacy is far from uniform; divergent training protocols and data sources cause individual TSFMs to exhibit highly variable performance across different forecasting tasks, domains, and horizons. Leveraging this complementary expertise by arbitrating existing TSFM outputs presents a compelling strategy, yet this remains a largely unexplored area of research. In this paper, we conduct a thorough examination of how different TSFMs exhibit specialized performance profiles across various forecasting settings, and how we can effectively leverage this behavior in arbitration between different time series models. We specifically analyze how factors such as model selection and forecast horizon distribution can influence the efficacy of arbitration strategies. Based on this analysis, we propose Synapse, a novel arbitration framework for TSFMs. Synapse is designed to dynamically leverage a pool of TSFMs, assign and adjust predictive weights based on their relative, context-dependent performance, and construct a robust forecast distribution by adaptively sampling from the output quantiles of constituent models. Experimental results demonstrate that Synapse consistently outperforms other popular ensembling techniques as well as individual TSFMs, demonstrating Synapse's efficacy in time series forecasting.

Booth is a reporter at TIME. Booth is a reporter at TIME. Jimmy Wales describes himself as a "pathological optimist." And yet, when the co-founder of Wikipedia spoke with TIME in October, he still seemed somewhat surprised that his online encyclopedia actually worked. "Wikipedia is very trusting, in a way that always seemed a bit crazy," Wales says.