--As large language models are increasingly deployed in sensitive environments, fingerprinting attacks pose significant privacy and security risks. We present a study of LLM fingerprinting from both offensive and defensive perspectives. Our attack methodology uses reinforcement learning to automatically optimize query selection, achieving better fingerprinting accuracy with only 3 queries compared to randomly selecting 3 queries from the same pool. Our defensive approach employs semantic-preserving output filtering through a secondary LLM to obfuscate model identity while maintaining semantic integrity. The defensive method reduces fingerprinting accuracy across tested models while preserving output quality. These contributions show the potential to improve fingerprinting tools capabilities while providing practical mitigation strategies against fingerprinting attacks. Large language models (LLMs) have become ubiquitous across industries, from customer service chatbots to code generation tools and content creation platforms. As organizations increasingly rely on these models for sensitive applications, the ability to identify which specific model generated a given text--known as LLM fingerprinting--has emerged as a critical security concern.

Since Knowledge Graphs are often incomplete, link prediction methods are adopted for predicting missing facts. Scalable embedding based solutions are mostly adopted for this purpose, however, they lack comprehensibility, which may be crucial in several domains. Explanation methods tackle this issue by identifying supporting knowledge explaining the predicted facts. Regretfully, evaluating/comparing quantitatively the resulting explanations is challenging as there is no standard evaluation protocol and overall benchmarking resource. We fill this important gap by proposing GRainsaCK, a reusable software resource that fully streamlines all the tasks involved in benchmarking explanations, i.e., from model training to evaluation of explanations along the same evaluation protocol. Moreover, GRainsaCK furthers modularity/extensibility by implementing the main components as functions that can be easily replaced. Finally, fostering its reuse, we provide extensive documentation including a tutorial.

The manual translation of unstructured team dialogue into the structured artifacts required for Information Technology (IT) project governance is a critical bottleneck in modern information systems management. We introduce DevNous, a Large Language Model-based (LLM) multi-agent expert system, to automate this unstructured-to-structured translation process. DevNous integrates directly into team chat environments, identifying actionable intents from informal dialogue and managing stateful, multi-turn workflows for core administrative tasks like automated task formalization and progress summary synthesis. To quantitatively evaluate the system, we introduce a new benchmark of 160 realistic, interactive conversational turns. The dataset was manually annotated with a multi-label ground truth and is publicly available. On this benchmark, DevNous achieves an exact match turn accuracy of 81.3\% and a multiset F1-Score of 0.845, providing strong evidence for its viability. The primary contributions of this work are twofold: (1) a validated architectural pattern for developing ambient administrative agents, and (2) the introduction of the first robust empirical baseline and public benchmark dataset for this challenging problem domain.

Language models have shown strong capabilities across a wide range of tasks in software engineering, such as code generation, yet they suffer from hallucinations. While hallucinations have been studied independently in natural language and code generation, their occurrence in tasks involving code changes which have a structurally complex and context-dependent format of code remains largely unexplored. This paper presents the first comprehensive analysis of hallucinations in two critical tasks involving code change to natural language generation: commit message generation and code review comment generation. We quantify the prevalence of hallucinations in recent language models and explore a range of metric-based approaches to automatically detect them. Our findings reveal that approximately 50\% of generated code reviews and 20\% of generated commit messages contain hallucinations. Whilst commonly used metrics are weak detectors on their own, combining multiple metrics substantially improves performance. Notably, model confidence and feature attribution metrics effectively contribute to hallucination detection, showing promise for inference-time detection.\footnote{All code and data will be released upon acceptance.

Accurate molecular property prediction is a critical challenge with wide-ranging applications in chemistry, materials science, and drug discovery. Molecular representation methods, including fingerprints and graph neural networks (GNNs), achieve state-of-the-art results by effectively deriving features from molecular structures. However, these methods often overlook decades of accumulated semantic and contextual knowledge. Recent advancements in large language models (LLMs) demonstrate remarkable reasoning abilities and prior knowledge across scientific domains, leading us to hypothesize that LLMs can generate rich molecular representations when guided to reason in multiple perspectives. To address these gaps, we propose $\text{M}^{2}$LLM, a multi-view framework that integrates three perspectives: the molecular structure view, the molecular task view, and the molecular rules view. These views are fused dynamically to adapt to task requirements, and experiments demonstrate that $\text{M}^{2}$LLM achieves state-of-the-art performance on multiple benchmarks across classification and regression tasks. Moreover, we demonstrate that representation derived from LLM achieves exceptional performance by leveraging two core functionalities: the generation of molecular embeddings through their encoding capabilities and the curation of molecular features through advanced reasoning processes.

Despite the rapid progress of Large Language Models (LLMs), their application in agriculture remains limited due to the lack of domain-specific models, curated datasets, and robust evaluation frameworks. To address these challenges, we propose AgriGPT, a domain-specialized LLM ecosystem for agricultural usage. At its core, we design a multi-agent scalable data engine that systematically compiles credible data sources into Agri-342K, a high-quality, standardized question-answer (QA) dataset. Trained on this dataset, AgriGPT supports a broad range of agricultural stakeholders, from practitioners to policy-makers. To enhance factual grounding, we employ Tri-RAG, a three-channel Retrieval-Augmented Generation framework combining dense retrieval, sparse retrieval, and multi-hop knowledge graph reasoning, thereby improving the LLM's reasoning reliability. For comprehensive evaluation, we introduce AgriBench-13K, a benchmark suite comprising 13 tasks with varying types and complexities. Experiments demonstrate that AgriGPT significantly outperforms general-purpose LLMs on both domain adaptation and reasoning. Beyond the model itself, AgriGPT represents a modular and extensible LLM ecosystem for agriculture, comprising structured data construction, retrieval-enhanced generation, and domain-specific evaluation. This work provides a generalizable framework for developing scientific and industry-specialized LLMs. All models, datasets, and code will be released to empower agricultural communities, especially in underserved regions, and to promote open, impactful research.

This review examines recent advances in Parameter-Efficient Fine-Tuning (PEFT), with a focus on Low-Rank Adaptation (LoRA), to optimize Retrieval-Augmented Generation (RAG) systems like Qwen3, DeepSeek, and Kimi. These systems face challenges in understanding and generating authentic Cantonese colloquial expressions due to limited annotated data and linguistic variability. The review evaluates the integration of LoRA within RAG frameworks, benchmarks PEFT methods for retrieval and generation accuracy, identify domain adaptation strategies under limited data, and compares fine-tuning techniques aimed at improving semantic fidelity under data-scarce conditions. A systematic analysis of recent studies employing diverse LoRA variants, synthetic data generation, user feedback integration, and adaptive parameter allocation was conducted to assess their impact on computational efficiency, retrieval precision, linguistic authenticity, and scalability. Findings reveal that dynamic and ensemble LoRA adaptations significantly reduce trainable parameters without sacrificing retrieval accuracy and generation quality in dialectal contexts. However, limitations remain in fully preserving fine-grained linguistic nuances, especially for low-resource settings like Cantonese. The integration of real-time user feedback and domain-specific data remains underdeveloped, limiting model adaptability and personalization. While selective parameter freezing and nonlinear adaptation methods offer better trade-offs between efficiency and accuracy, their robustness at scale remains an open challenge. This review highlights the promise of PEFT-enhanced RAG systems for domain-specific language tasks and calls for future work targeting dialectal authenticity, dynamic adaptation, and scalable fine-tuning pipelines.

As the social implementation of AI has been steadily progressing, research and development related to AI security has also been increasing. However, existing studies have been limited to organizing related techniques, attacks, defenses, and risks in terms of specific domains or AI elements. Thus, it extremely difficult to understand the relationships among them and how negative impacts on stakeholders are brought about. In this paper, we argue that the knowledge, technologies, and social impacts related to AI security should be holistically organized to help understand relationships among them. To this end, we first develop an AI security map that holistically organizes interrelationships among elements related to AI security as well as negative impacts on information systems and stakeholders. This map consists of the two aspects, namely the information system aspect (ISA) and the external influence aspect (EIA). The elements that AI should fulfill within information systems are classified under the ISA. The EIA includes elements that affect stakeholders as a result of AI being attacked or misused. For each element, corresponding negative impacts are identified. By referring to the AI security map, one can understand the potential negative impacts, along with their causes and countermeasures. Additionally, our map helps clarify how the negative impacts on AI-based systems relate to those on stakeholders. We show some findings newly obtained by referring to our map. We also provide several recommendations and open problems to guide future AI security communities.

Developers insert logging statements in source code to capture relevant runtime information essential for maintenance and debugging activities. Log level choice is an integral, yet tricky part of the logging activity as it controls log verbosity and therefore influences systems' observability and performance. Recent advances in ML-based log level prediction have leveraged large language models (LLMs) to propose log level predictors (LLPs) that demonstrated promising performance improvements (AUC between 0.64 and 0.8). Nevertheless, current LLM-based LLPs rely on randomly selected in-context examples, overlooking the structure and the diverse logging practices within modern software projects. In this paper, we propose OmniLLP, a novel LLP enhancement framework that clusters source files based on (1) semantic similarity reflecting the code's functional purpose, and (2) developer ownership cohesion. By retrieving in-context learning examples exclusively from these semantic and ownership aware clusters, we aim to provide more coherent prompts to LLPs leveraging LLMs, thereby improving their predictive accuracy. Our results show that both semantic and ownership-aware clusterings statistically significantly improve the accuracy (by up to 8\% AUC) of the evaluated LLM-based LLPs compared to random predictors (i.e., leveraging randomly selected in-context examples from the whole project). Additionally, our approach that combines the semantic and ownership signal for in-context prediction achieves an impressive 0.88 to 0.96 AUC across our evaluated projects. Our findings highlight the value of integrating software engineering-specific context, such as code semantic and developer ownership signals into LLM-LLPs, offering developers a more accurate, contextually-aware approach to logging and therefore, enhancing system maintainability and observability.

Ontology alignment plays a crucial role in integrating diverse data sources across domains. There is a large plethora of systems that tackle the ontology alignment problem, yet challenges persist in producing highly quality correspondences among a set of input ontologies. Human-in-the-loop during the alignment process is essential in applications requiring very accurate mappings. User involvement is, however, expensive when dealing with large ontologies. In this paper, we explore the feasibility of using Large Language Models (LLM) as an alternative to the domain expert. The use of the LLM focuses only on the validation of the subset of correspondences where an ontology alignment system is very uncertain. We have conducted an extensive evaluation over several matching tasks of the Ontology Alignment Evaluation Initiative (OAEI), analysing the performance of several state-of-the-art LLMs using different ontology-driven prompt templates. The LLM results are also compared against simulated Oracles with variable error rates.