Russian forces attacked Ukraine's capital, Kyiv, early on Monday, wounding five people and damaging a residential building, according to the head of the city's military administration, Tymur Tkachenko. A Russian drone hit a Ukrainian bus carrying 39 evacuees in the eastern Sumy region, near Ukraine's border with Russia, on Sunday, killing three people and wounding 19 others, according to the regional governor. Two others were killed in a landmine explosion in Sumy's Esman community on Saturday, while two more were killed in Russian attacks on the front-line Donetsk region, according to officials, taking the death toll from attacks across Ukraine on that day to at least six. Ukraine's forces also launched drone attacks at Russia on Sunday, with the governor of the Leningrad region reporting that at least 10 Ukrainian unmanned aircraft were downed over the areas surrounding the city of St Petersburg. Falling debris injured a woman, Governor Alexander Drozdenko said.

As large language models (LLMs) increasingly integrate native code interpreters, they enable powerful real-time execution capabilities, substantially expanding their utility. However, such integrations introduce potential system-level cybersecurity threats, fundamentally different from prompt-based vulnerabilities. To systematically evaluate these interpreter-specific risks, we propose CIRCLE (Code-Interpreter Resilience Check for LLM Exploits), a simple benchmark comprising 1,260 prompts targeting CPU, memory, and disk resource exhaustion. Each risk category includes explicitly malicious ("direct") and plausibly benign ("indirect") prompt variants. Our automated evaluation framework assesses not only whether LLMs refuse or generates risky code, but also executes the generated code within the interpreter environment to evaluate code correctness, simplifications made by the LLM to make the code safe, or execution timeouts. Evaluating 7 commercially available models from OpenAI and Google, we uncover significant and inconsistent vulnerabilities. For instance, evaluations show substantial disparities even within providers - OpenAI's o4-mini correctly refuses risky requests at 7.1%, notably higher rates compared to GPT-4.1 at 0.5%. Results particularly underscore that indirect, socially-engineered prompts substantially weaken model defenses. This highlights an urgent need for interpreter-specific cybersecurity benchmarks, dedicated mitigation tools (e.g., guardrails), and clear industry standards to guide safe and responsible deployment of LLM interpreter integrations. The benchmark dataset and evaluation code are publicly released to foster further research.

Many recent papers have studied the development of superforecaster-level event forecasting LLMs. While methodological problems with early studies cast doubt on the use of LLMs for event forecasting, recent studies with improved evaluation methods have shown that state-of-the-art LLMs are gradually reaching superforecaster-level performance, and reinforcement learning has also been reported to improve future forecasting. Additionally, the unprecedented success of recent reasoning models and Deep Research-style models suggests that technology capable of greatly improving forecasting performance has been developed. Therefore, based on these positive recent trends, we argue that the time is ripe for research on large-scale training of superforecaster-level event forecasting LLMs. We discuss two key research directions: training methods and data acquisition. For training, we first introduce three difficulties of LLM-based event forecasting training: noisiness-sparsity, knowledge cut-off, and simple reward structure problems. Then, we present related ideas to mitigate these problems: hypothetical event Bayesian networks, utilizing poorly-recalled and counterfactual events, and auxiliary reward signals. For data, we propose aggressive use of market, public, and crawling datasets to enable large-scale training and evaluation. Finally, we explain how these technical advances could enable AI to provide predictive intelligence to society in broader areas. This position paper presents promising specific paths and considerations for getting closer to superforecaster-level AI technology, aiming to call for researchers' interest in these directions.

Large language models (LLMs) are increasingly adapted to downstream tasks via reinforcement learning (RL) methods like Group Relative Policy Optimization (GRPO), which often require thousands of rollouts to learn new tasks. We argue that the interpretable nature of language can often provide a much richer learning medium for LLMs, compared with policy gradients derived from sparse, scalar rewards. To test this, we introduce GEPA (Genetic-Pareto), a prompt optimizer that thoroughly incorporates natural language reflection to learn high-level rules from trial and error. Given any AI system containing one or more LLM prompts, GEPA samples system-level trajectories (e.g., reasoning, tool calls, and tool outputs) and reflects on them in natural language to diagnose problems, propose and test prompt updates, and combine complementary lessons from the Pareto frontier of its own attempts. As a result of GEPA's design, it can often turn even just a few rollouts into a large quality gain. Across four tasks, GEPA outperforms GRPO by 10% on average and by up to 20%, while using up to 35x fewer rollouts. GEPA also outperforms the leading prompt optimizer, MIPROv2, by over 10% across two LLMs, and demonstrates promising results as an inference-time search strategy for code optimization.

Capturing latent variations ( "contexts") is key to deploying reinforcement-learning (RL) agents beyond their training regime. We recast context-based RL as a dual inference-control problem and formally characterize two properties and their hierarchy: observation sufficiency (preserving all predictive information) and control sufficiency (retaining decision-making relevant information). Exploiting this dichotomy, we derive a contextual evidence lower bound(ELBO)-style objective that cleanly separates representation learning from policy learning and optimizes it with Bottlenecked Contextual Policy Optimization (BCPO), an algorithm that places a variational information-bottleneck encoder in front of any off-policy policy learner. On standard continuous-control benchmarks with shifting physical parameters, BCPO matches or surpasses other baselines while using fewer samples and retaining performance far outside the training regime.

Large Language Models (LLMs) have demonstrated remarkable capabilities across a wide range of instruction-following tasks, yet their grasp of nuanced social science concepts remains underexplored. This paper examines whether LLMs can identify and classify fine-grained forms of populism, a complex and contested concept in both academic and media debates. To this end, we curate and release novel datasets specifically designed to capture populist discourse. We evaluate a range of pre-trained (large) language models, both open-weight and proprietary, across multiple prompting paradigms. Our analysis reveals notable variation in performance, highlighting the limitations of LLMs in detecting populist discourse. We find that a fine-tuned RoBERTa classifier vastly outperforms all new-era instruction-tuned LLMs, unless fine-tuned. Additionally, we apply our best-performing model to analyze campaign speeches by Donald Trump, extracting valuable insights into his strategic use of populist rhetoric. Finally, we assess the generalizability of these models by benchmarking them on campaign speeches by European politicians, offering a lens into cross-context transferability in political discourse analysis. In this setting, we find that instruction-tuned LLMs exhibit greater robustness on out-of-domain data.

Static benchmarks fail to capture LLM vulnerabilities emerging through community experimentation in online forums. We present PrompTrend, a system that collects vulnerability data across platforms and evaluates them using multidimensional scoring, with an architecture designed for scalable monitoring. Cross-sectional analysis of 198 vulnerabilities collected from online communities over a five-month period (January-May 2025) and tested on nine commercial models reveals that advanced capabilities correlate with increased vulnerability in some architectures, psychological attacks significantly outperform technical exploits, and platform dynamics shape attack effectiveness with measurable model-specific patterns. The PrompTrend Vulnerability Assessment Framework achieves 78% classification accuracy while revealing limited cross-model transferability, demonstrating that effective LLM security requires comprehensive socio-technical monitoring beyond traditional periodic assessment. Our findings challenge the assumption that capability advancement improves security and establish community-driven psychological manipulation as the dominant threat vector for current language models.

The increasing use of Large Language Models (LLMs) in a large variety of domains has sparked worries about how easily they can perpetuate stereotypes and contribute to the generation of biased content. With a focus on gender and professional bias, this work examines in which manner LLMs shape responses to ungendered prompts, contributing to biased outputs. This analysis uses a structured experimental method, giving different prompts involving three different professional job combinations, which are also characterized by a hierarchical relationship. This study uses Italian, a language with extensive grammatical gender differences, to highlight potential limitations in current LLMs' ability to generate objective text in non-English languages. Two popular LLM-based chatbots are examined, namely OpenAI ChatGPT (gpt-4o-mini) and Google Gemini (gemini-1.5-flash). Through APIs, we collected a range of 3600 responses. The results highlight how content generated by LLMs can perpetuate stereotypes. For example, Gemini associated 100% (ChatGPT 97%) of 'she' pronouns to the 'assistant' rather than the 'manager'. The presence of bias in AI-generated text can have significant implications in many fields, such as in the workplaces or in job selections, raising ethical concerns about its use. Understanding these risks is pivotal to developing mitigation strategies and assuring that AI-based systems do not increase social inequalities, but rather contribute to more equitable outcomes. Future research directions include expanding the study to additional chatbots or languages, refining prompt engineering methods or further exploiting a larger experimental base.

Interpretation of topics is crucial for their downstream applications. State-of-the-art evaluation measures of topic quality such as coherence and word intrusion do not measure how much a topic facilitates the exploration of a corpus. To design evaluation measures grounded on a task, and a population of users, we do user studies to understand how users interpret topics. We propose constructs of topic quality and ask users to assess them in the context of a topic and provide rationale behind evaluations. We use reflexive thematic analysis to identify themes of topic interpretations from rationales. Users interpret topics based on availability and representativeness heuristics rather than probability. We propose a theory of topic interpretation based on the anchoring-and-adjustment heuristic: users anchor on salient words and make semantic adjustments to arrive at an interpretation. Topic interpretation can be viewed as making a judgment under uncertainty by an ecologically rational user, and hence cognitive biases aware user models and evaluation frameworks are needed.

Ensuring privacy in large-scale open datasets is increasingly challenging under regulations such as the General Data Protection Regulation (GDPR). While differential privacy (DP) provides strong theoretical guarantees, it primarily focuses on noise injection during model training, neglecting privacy preservation at the data publishing stage. Existing privacy-preserving data publishing (PPDP) approaches struggle to balance privacy and utility, particularly when data publishers and users are distinct entities. To address this gap, we focus on the graph recovery problem and propose a novel privacy-preserving estimation framework for open graph data, leveraging Gaussian DP (GDP) with a structured noise-injection mechanism. Unlike traditional methods that perturb gradients or model updates, our approach ensures unbiased graph structure recovery while enforcing DP at the data publishing stage. Moreover, we provide theoretical guarantees on estimation accuracy and extend our method to discrete-variable graphs, a setting often overlooked in DP research. Experimental results in graph learning demonstrate robust performance, offering a viable solution for privacy-conscious graph analysis.