Different transformer architectures implement identical linguistic computations via distinct connectivity patterns, yielding model imprinted ``computational fingerprints'' detectable through spectral analysis. Using graph signal processing on attention induced token graphs, we track changes in algebraic connectivity (Fiedler value, $Δλ_2$) under voice alternation across 20 languages and three model families, with a prespecified early window (layers 2--5). Our analysis uncovers clear architectural signatures: Phi-3-Mini shows a dramatic English specific early layer disruption ($\overline{Δλ_2}_{[2,5]}\!\approx\!-0.446$) while effects in 19 other languages are minimal, consistent with public documentation that positions the model primarily for English use. Qwen2.5-7B displays small, distributed shifts that are largest for morphologically rich languages, and LLaMA-3.2-1B exhibits systematic but muted responses. These spectral signatures correlate strongly with behavioral differences (Phi-3: $r=-0.976$) and are modulated by targeted attention head ablations, linking the effect to early attention structure and confirming functional relevance. Taken together, the findings are consistent with the view that training emphasis can leave detectable computational imprints: specialized processing strategies that manifest as measurable connectivity patterns during syntactic transformations. Beyond voice alternation, the framework differentiates reasoning modes, indicating utility as a simple, training free diagnostic for revealing architectural biases and supporting model reliability analysis.

The advent of Large Language Models has revolutionized tasks across domains, including the automation of legal document analysis, a critical component of modern contract management systems. This paper presents a comprehensive implementation of LLM-enhanced metadata extraction for contract review, focusing on the automatic detection and annotation of salient legal clauses. Leveraging both the publicly available Contract Understanding Atticus Dataset (CUAD) and proprietary contract datasets, our work demonstrates the integration of advanced LLM methodologies with practical applications. We identify three pivotal elements for optimizing metadata extraction: robust text conversion, strategic chunk selection, and advanced LLM-specific techniques, including Chain of Thought (CoT) prompting and structured tool calling. The results from our experiments highlight the substantial improvements in clause identification accuracy and efficiency. Our approach shows promise in reducing the time and cost associated with contract review while maintaining high accuracy in legal clause identification. The results suggest that carefully optimized LLM systems could serve as valuable tools for legal professionals, potentially increasing access to efficient contract review services for organizations of all sizes.

Neural activity forecasting is central to understanding neural systems and enabling closed-loop control. While deep learning has recently advanced the state-of-the-art in the time series forecasting literature, its application to neural activity forecasting remains limited. To bridge this gap, we systematically evaluated eight probabilistic deep learning models, including two foundation models, that have demonstrated strong performance on general forecasting benchmarks. We compared them against four classical statistical models and two baseline methods on spontaneous neural activity recorded from mouse cortex via widefield imaging. Across prediction horizons, several deep learning models consistently outperformed classical approaches, with the best model producing informative forecasts up to 1.5 seconds into the future. Our findings point toward future control applications and open new avenues for probing the intrinsic temporal structure of neural activity.

Rigorous security-focused evaluation of large language model (LLM) agents is imperative for establishing trust in their safe deployment throughout the software development lifecycle. However, existing benchmarks largely rely on synthetic challenges or simplified vulnerability datasets that fail to capture the complexity and ambiguity encountered by security engineers in practice. We introduce SEC-bench, the first fully automated benchmarking framework for evaluating LLM agents on authentic security engineering tasks. SEC-bench employs a novel multi-agent scaffold that automatically constructs code repositories with harnesses, reproduces vulnerabilities in isolated environments, and generates gold patches for reliable evaluation. Our framework automatically creates high-quality software vulnerability datasets with reproducible artifacts at a cost of only $0.87 per instance. Using SEC-bench, we implement two critical software security tasks to rigorously evaluate LLM agents' capabilities: proof-of-concept (PoC) generation and vulnerability patching. A comprehensive evaluation of state-of-the-art LLM code agents reveals significant performance gaps, achieving at most 18.0% success in PoC generation and 34.0% in vulnerability patching on our complete dataset. These results highlight the crucial steps needed toward developing LLM agents that are more practical, intelligent, and autonomous for security engineering.

This pilot study traces students' reflections on the use of AI in a 13-week foundational design course enrolling over 500 first-year engineering and architecture students at the Singapore University of Technology and Design. The course was an AI-enhanced design course, with several interventions to equip students with AI based design skills. Students were required to reflect on whether the technology was used as a tool (instrumental assistant), a teammate (collaborative partner), or neither (deliberate non-use). By foregrounding this three-way lens, students learned to use AI for innovation rather than just automation and to reflect on agency, ethics, and context rather than on prompt crafting alone. Evidence stems from coursework artefacts: thirteen structured reflection spreadsheets and eight illustrated briefs submitted, combined with notes of teachers and researchers. Qualitative coding of these materials reveals shared practices brought about through the inclusion of Gen-AI, including accelerated prototyping, rapid skill acquisition, iterative prompt refinement, purposeful "switch-offs" during user research, and emergent routines for recognizing hallucinations. Unexpectedly, students not only harnessed Gen-AI for speed but (enabled by the tool-teammate-neither triage) also learned to reject its outputs, invent their own hallucination fire-drills, and divert the reclaimed hours into deeper user research, thereby transforming efficiency into innovation. The implications of the approach we explore shows that: we can transform AI uptake into an assessable design habit; that rewarding selective non-use cultivates hallucination-aware workflows; and, practically, that a coordinated bundle of tool access, reflection, role tagging, and public recognition through competition awards allows AI based innovation in education to scale without compromising accountability.

LLMs have recently demonstrated strong potential in simulating online shopper behavior. Prior work has improved action prediction by applying SFT on action traces with LLM-generated rationales, and by leveraging RL to further enhance reasoning capabilities. Despite these advances, current approaches rely on text-based inputs and overlook the essential role of visual perception in shaping human decision-making during web GUI interactions. In this paper, we investigate the integration of visual information, specifically webpage screenshots, into behavior simulation via VLMs, leveraging OPeRA dataset. By grounding agent decision-making in both textual and visual modalities, we aim to narrow the gap between synthetic agents and real-world users, thereby enabling more cognitively aligned simulations of online shopping behavior. Specifically, we employ SFT for joint action prediction and rationale generation, conditioning on the full interaction context, which comprises action history, past HTML observations, and the current webpage screenshot. To further enhance reasoning capabilities, we integrate RL with a hierarchical reward structure, scaled by a difficulty-aware factor that prioritizes challenging decision points. Empirically, our studies show that incorporating visual grounding yields substantial gains: the combination of text and image inputs improves exact match accuracy by more than 6% over text-only inputs. These results indicate that multi-modal grounding not only boosts predictive accuracy but also enhances simulation fidelity in visually complex environments, which captures nuances of human attention and decision-making that text-only agents often miss. Finally, we revisit the design space of behavior simulation frameworks, identify key methodological limitations, and propose future research directions toward building efficient and effective human behavior simulators.

We introduce CosmoCore, a neuroscience-inspired reinforcement learning (RL) architecture that integrates affective signals to enhance code generation in large language models (LLMs). Motivated by human and animal learning where embarrassment from mistakes drives rapid correction, as observed in training a puppy to avoid repeating errors after a single scolding CosmoCore tags code generation trajectories with valence and surprise using a lightweight multi-layer perceptron (MLP). High-negative valence (cringe) episodes, such as buggy code outputs, are prioritized in a Dream Queue for five-fold replay during off-policy updates, while low-surprise successes are pruned to prevent overconfidence and buffer bloat. Evaluated on code generation benchmarks like HumanEval and BigCodeBench, alongside simulations with a custom data pipeline environment, CosmoCore reduces hallucinated code (e.g., syntax errors or logical bugs) by 48\% and accelerates self-correction by 45\%. Local experiments using Hugging Face models in a PySpark environment validate these gains, with code snippets provided for replication. Ablations confirm valence tagging boosts curiosity in exploration, and pruning mitigates inefficiency. This framework extends RL from human feedback (RLHF) for more emotionally aware code assistants, with applications in IDEs and data pipelines. Code and the custom mini-world simulation are released.

Misuse of LLM-generated texts Modern LLMs are increasingly capable of generating highly fluent humanlike texts and adaptive to multiple dialects across genres [21]. While this progress unlocks many opportunities towards resolving daily life challenges, it also introduces risks in distinguishing between human-and machine-generated texts [22]. Undetected contents can cause serious cyber threats including misinformation, academic dishonesty, and even more aggressive consequences with phishing, smishing, and social engineering, where convincing texts are often used to manipulate individuals and organizations [23, 24]. For instance, the Anti-Phishing Working Group (APWG) have indicated in their phishing activity trends report that 932,923 phishing attacks were recorded worldwide in the third quarter of 2024, highlighting a significant increase in smishing by 22% during the same period. APWG also reported that social media platforms were the most targeted sector, representing 30.5% of all phishing attacks [24]. Furthermore, according to the FBI Internet Crime Report, losses due to cybercrimes in the United States have exceeded $12.5 billion in 2023 [25]. These examples illustrate how malicious manoeuvres can exploit machine-generated texts to scale deception. Hence, robust methods for detecting LLM-generated texts are urgently needed particularly with linguistically sensitive scripts such as Arabic, where formal, legal, and religious texts require extra reliability.

Enhancing large vision-language models (LVLMs) with visual slow-thinking reasoning is crucial for solving complex multimodal tasks. However, since LVLMs are mainly trained with vision-language alignment, it is difficult to adopt on-policy reinforcement learning (RL) to develop the slow thinking ability because the rollout space is restricted by its initial abilities. Off-policy RL offers a way to go beyond the current policy, but directly distilling trajectories from external models may cause visual hallucinations due to mismatched visual perception abilities across models. To address these issues, this paper proposes SOPHIA, a simple and scalable Semi-Off-Policy RL for vision-language slow-tHInking reAsoning. SOPHIA builds a semi-off-policy behavior model by combining on-policy visual understanding from a trainable LVLM with off-policy slow-thinking reasoning from a language model, assigns outcome-based rewards to reasoning, and propagates visual rewards backward. Then LVLM learns slow-thinking reasoning ability from the obtained reasoning trajectories using propagated rewards via off-policy RL algorithms. Extensive experiments with InternVL2.5 and InternVL3.0 with 8B and 38B sizes show the effectiveness of SOPHIA. Notably, SOPHIA improves InternVL3.0-38B by 8.50% in average, reaching state-of-the-art performance among open-source LVLMs on multiple multimodal reasoning benchmarks, and even outperforms some closed-source models (e.g., GPT-4.1) on the challenging MathVision and OlympiadBench, achieving 49.08% and 49.95% pass@1 accuracy, respectively. Analysis shows SOPHIA outperforms supervised fine-tuning and direct on-policy RL methods, offering a better policy initialization for further on-policy training.

Intrusion Detection and Prevention Systems (IDS/IPS) in large enterprises can generate hundreds of thousands of alerts per hour, overwhelming analysts with logs requiring rapidly evolving expertise. Conventional machine-learning detectors reduce alert volume but still yield many false positives, while standard Retrieval-Augmented Generation (RAG) pipelines often retrieve irrelevant context and fail to justify predictions. We present CyberRAG, a modular agent-based RAG framework that delivers real-time classification, explanation, and structured reporting for cyber-attacks. A central LLM agent orchestrates: (i) fine-tuned classifiers specialized by attack family; (ii) tool adapters for enrichment and alerting; and (iii) an iterative retrieval-and-reason loop that queries a domain-specific knowledge base until evidence is relevant and self-consistent. Unlike traditional RAG, CyberRAG adopts an agentic design that enables dynamic control flow and adaptive reasoning. This architecture autonomously refines threat labels and natural-language justifications, reducing false positives and enhancing interpretability. It is also extensible: new attack types can be supported by adding classifiers without retraining the core agent. CyberRAG was evaluated on SQL Injection, XSS, and SSTI, achieving over 94\% accuracy per class and a final classification accuracy of 94.92\% through semantic orchestration. Generated explanations reached 0.94 in BERTScore and 4.9/5 in GPT-4-based expert evaluation, with robustness preserved against adversarial and unseen payloads. These results show that agentic, specialist-oriented RAG can combine high detection accuracy with trustworthy, SOC-ready prose, offering a flexible path toward partially automated cyber-defense workflows.