Recent advances in Large Language Models (LLM) have led to a new class of autonomous agents, renewing and expanding interest in the area. LLM-powered Multiagent Systems (MAS) have thus emerged, both for assistive and simulation purposes, yet tools for realistic human behavior simulation -- with its distinctive challenges and opportunities -- remain underdeveloped. Existing MAS libraries and tools lack fine-grained persona specifications, population sampling facilities, experimentation support, and integrated validation, among other key capabilities, limiting their utility for behavioral studies, social simulation, and related applications. To address these deficiencies, in this work we introduce TinyTroupe, a simulation toolkit enabling detailed persona definitions (e.g., nationality, age, occupation, personality, beliefs, behaviors) and programmatic control via numerous LLM-driven mechanisms. This allows for the concise formulation of behavioral problems of practical interest, either at the individual or group level, and provides effective means for their solution. TinyTroupe's components are presented using representative working examples, such as brainstorming and market research sessions, thereby simultaneously clarifying their purpose and demonstrating their usefulness. Quantitative and qualitative evaluations of selected aspects are also provided, highlighting possibilities, limitations, and trade-offs. The approach, though realized as a specific Python implementation, is meant as a novel conceptual contribution, which can be partially or fully incorporated in other contexts. The library is available as open source at https://github.com/microsoft/tinytroupe.

Hacker forums provide critical early warning signals for emerging cybersecurity threats, but extracting actionable intelligence from their unstructured and noisy content remains a significant challenge. This paper presents an unsupervised framework that automatically detects, clusters, and prioritizes security events discussed across hacker forum posts. Our approach leverages Transformer-based embeddings fine-tuned with contrastive learning to group related discussions into distinct security event clusters, identifying incidents like zero-day disclosures or malware releases without relying on predefined keywords. The framework incorporates a daily ranking mechanism that prioritizes identified events using quantifiable metrics reflecting timeliness, source credibility, information completeness, and relevance. Experimental evaluation on real-world hacker forum data demonstrates that our method effectively reduces noise and surfaces high-priority threats, enabling security analysts to mount proactive responses. By transforming disparate hacker forum discussions into structured, actionable intelligence, our work addresses fundamental challenges in automated threat detection and analysis.

Large language models exhibit cultural biases and limited cross-cultural understanding capabilities, particularly when serving diverse global user populations. We propose MCEval, a novel multilingual evaluation framework that employs dynamic cultural question construction and enables causal analysis through Counterfactual Rephrasing and Confounder Rephrasing. Our comprehensive evaluation spans 13 cultures and 13 languages, systematically assessing both cultural awareness and cultural bias across different linguistic scenarios. The framework provides 39,897 cultural awareness instances and 17,940 cultural bias instances. Experimental results reveal performance disparities across different linguistic scenarios, demonstrating that optimal cultural performance is not only linked to training data distribution, but also is related to language-culture alignment. The evaluation results also expose the fairness issue, where approaches appearing successful in the English scenario create substantial disadvantages. MCEval represents the first comprehensive multilingual cultural evaluation framework that provides deeper insights into LLMs' cultural understanding.

Programming-by-prompting with generative AI offers a new paradigm for end-user programming, shifting the focus from syntactic fluency to semantic intent. This shift holds particular promise for non-programmers such as educators, who can describe instructional goals in natural language to generate interactive learning content. Yet in bypassing direct code authoring, many of programming's core affordances - such as traceability, stepwise refinement, and behavioral testing - are lost. We propose the Chain-of-Abstractions (CoA) framework as a way to recover these affordances while preserving the expressive flexibility of natural language. CoA decomposes the synthesis process into a sequence of cognitively meaningful, task-aligned representations that function as checkpoints for specification, inspection, and refinement. We instantiate this approach in SimStep, an authoring environment for teachers that scaffolds simulation creation through four intermediate abstractions: Concept Graph, Scenario Graph, Learning Goal Graph, and UI Interaction Graph. To address ambiguities and misalignments, SimStep includes an inverse correction process that surfaces in-filled model assumptions and enables targeted revision without requiring users to manipulate code. Evaluations with educators show that CoA enables greater authoring control and interpretability in programming-by-prompting workflows.

Large reasoning models (LRMs) like OpenAI o1 and DeepSeek R1 have demonstrated impressive performance on complex reasoning tasks like mathematics and programming with long Chain-of-Thought (CoT) reasoning sequences (slow-thinking), compared with traditional large language models (fast-thinking). However, these reasoning models also face a huge challenge that generating unnecessarily lengthy and redundant reasoning chains even for trivial questions. This phenomenon leads to a significant waste of inference resources, increases the response time for simple queries, and hinders the practical application of LRMs in real-world products. To this end, it is crucial to shorten lengthy reasoning chains and learn adaptive reasoning between fast and slow thinking based on input difficulty. In this survey, we provide a comprehensive overview of recent progress in concise and adaptive thinking for efficient reasoning of LRMs, including methodologies, benchmarks, and challenges for future exploration. We hope this survey can help researchers quickly understand the landscape of this field and inspire novel adaptive thinking ideas to facilitate better usage of LRMs.

We use generative agents powered by large language models (LLMs) to simulate a social network in a shared residential building, driving the temperature decisions for a central heating system. Agents, divided into Family Members and Representatives, consider personal preferences, personal traits, connections, and weather conditions. Daily simulations involve family-level consensus followed by building-wide decisions among representatives. We tested three personality traits distributions (positive, mixed, and negative) and found that positive traits correlate with higher happiness and stronger friendships. Temperature preferences, assertiveness, and selflessness have a significant impact on happiness and decisions. This work demonstrates how LLM-driven agents can help simulate nuanced human behavior where complex real-life human simulations are di fficult to set. Introduction Social network simulations are widely utilized to model the interactions between people, often relying on agent-based modeling to represent the relationship between people and their environment. In these simulations, the actions of the agents are selected from a predefined set of rules specified by the modeler. While this rule-based approach allows for a clear definition of the decision process and provides control over the outcomes, it also introduces a limitation. The predefined rules may not be able to model various dimensions of human behavior, such as irrational decision-making, and restrict the knowledge of agents to what is encoded by the modeler. Large language models (LLMs) are trained on a vast amount of data, mostly obtained from web pages (Wang et al., 2024). Learning from this human-generated data allows the models to have a level of real-world knowledge and reduces the amount of external information that is required to be given to perform various tasks. Moreover, with their abilities such as reasoning and role-playing, LLMs have previously been shown to have significant capabilities of simulating human-like behavior. Generative agents, as introduced in Park et al. (2023), rely on LLMs to generate agent behaviors, based on agent-specific memory about the agent's identity, interactions with the other agents, and the environment.

The Retrieval-Augmented Generation (RAG) systems' performance on Thai legal question answering is still limited, especially for questions requiring extensive, complex legal reasoning. To address these limitations, we introduce an approach aligning LLMs toward improved law citation accuracy and better response quality using Group-Relative Policy Optimization (GRPO). Our approach leverages BGE-M3 embeddings as a cost-efficient semantic-similarity reward, significantly reducing computational expenses up to 2.5x compared to large language model judges. Experiments on the NitiBench benchmark demonstrate substantial improvements: GRPO achieves up to 90% citation-F1 gains from the base model and a 31% increase in joint quality metrics over instruction tuning. Crucially, our method shows enhanced robustness on complex legal reasoning tasks compared to instruction tuning, providing an effective and resource-efficient solution for enhancing Thai legal LLMs.

Personal service robots are deployed to support daily living in domestic environments, particularly for elderly and individuals requiring assistance. These robots must perceive complex and dynamic surroundings, understand tasks, and execute context-appropriate actions. However, current systems rely on proprietary, hard-coded solutions tied to specific hardware and software, resulting in siloed implementations that are difficult to adapt and scale across platforms. Ontologies and Knowledge Graphs (KGs) offer a solution to enable interoperability across systems, through structured and standardized representations of knowledge and reasoning. However, symbolic systems such as KGs and ontologies struggle with raw and noisy sensory input. In contrast, multimodal language models are well suited for interpreting input such as images and natural language, but often lack transparency, consistency, and knowledge grounding. In this work, we propose a neurosymbolic framework that combines the perceptual strengths of multimodal language models with the structured representations provided by KGs and ontologies, with the aim of supporting interoperability in robotic applications. Our approach generates ontology-compliant KGs that can inform robot behavior in a platform-independent manner. We evaluated this framework by integrating robot perception data, ontologies, and five multimodal models (three LLaMA and two GPT models), using different modes of neural-symbolic interaction. We assess the consistency and effectiveness of the generated KGs across multiple runs and configurations, and perform statistical analyzes to evaluate performance. Results show that GPT-o1 and LLaMA 4 Maverick consistently outperform other models. However, our findings also indicate that newer models do not guarantee better results, highlighting the critical role of the integration strategy in generating ontology-compliant KGs.

Deploying transformer-based neural networks on resource-constrained edge devices presents a significant challenge. This challenge is often addressed through various techniques, such as low-rank approximation and mixed-precision quantization. In this work, we introduce Mixed Low-Rank and Quantization (MLoRQ), a novel method that integrates both techniques. MLoRQ employs a two-stage optimization process to determine optimal bit-width and rank assignments for each layer, adhering to predefined memory constraints. This process includes: (i) an intra-layer optimization that identifies potentially optimal compression solutions out of all low-rank and quantization combinations; (ii) an inter-layer optimization that assigns bit-width precision and rank to each layer while ensuring the memory constraint is met. An optional final step applies a sequential optimization process using a modified adaptive rounding technique to mitigate compression-induced errors in joint low-rank approximation and quantization. The method is compatible and can be seamlessly integrated with most existing quantization algorithms. MLoRQ shows state-of-the-art results with up to 15\% performance improvement, evaluated on Vision Transformers for image classification, object detection, and instance segmentation tasks.

Artificial intelligence (AI) has made remarkable progress in recent years, yet its rapid expansion brings overlooked environmental and ethical challenges. This review explores four critical areas where AI's impact extends beyond performance: energy consumption, electronic waste (e-waste), inequality in compute access, and the hidden energy burden of cybersecurity systems. Drawing from recent studies and institutional reports, the paper highlights systemic issues such as high emissions from model training, rising hardware turnover, global infrastructure disparities, and the energy demands of securing AI. By connecting these concerns, the review contributes to Responsible AI discourse by identifying key research gaps and advocating for sustainable, transparent, and equitable development practices. Ultimately, it argues that AI's progress must align with ethical responsibility and environmental stewardship to ensure a more inclusive and sustainable technological future.