Empirical Study

Large Language Models (LLMs) are increasingly used to generate summaries of software bug reports, including sections such as Steps-to-Reproduce (S2R), Actual Behavior (AB), and Expected Behavior (EB). However, these models frequently produce hallucinations that can be convincing but unsupported by the source report. This can mislead developers and reduce trust in automated maintenance tools. Existing hallucination detection approaches typically evaluate outputs at the full-response level and do not consider the structure of technical documents. An initial exploratory study on 80 structured bug report summaries found that approximately 47.9% contained missing information, while 12.3% included fabricated content, highlighting the need for systematic hallucination analysis in bug report summarization. In this work, we empirically investigate hallucinations in LLM-generated bug report summaries from a section-aware perspective. Using the BugsRepo dataset, derived from Mozilla OSS projects, we introduce controlled synthetic hallucination injection to construct a benchmark for training and evaluation. We propose a section-aware hallucination detection approach that jointly predicts whether a summary contains hallucinated content, identifies affected sections, and classifies hallucination types. Experimental results across multiple pretrained language models show that the proposed approach achieves strong performance across all tasks, with the best model obtaining 0.89 report-level Macro-F1, 0.83 section-level Macro-F1, and 0.84 hallucination-type Macro-F1. We further analyze common hallucination patterns and model failure modes to better understand limitations of current LLM-generated bug report summaries. The findings highlight the importance of section-aware hallucination analysis for improving the reliability of LLM-assisted bug report summarization in software maintenance workflows.

AI-assisted code review tools typically operate as generic "expert reviewer" agents, producing homogeneous findings regardless of the analysis type needed. We present a system that constrains AI reviewer behavior through philosophical dispositions -- coherent personality lenses grounded in specific epistemological traditions (Pyrrhonist Skepticism, Navya-Ny=aya logic, Diogenes' Cynicism, Confucian relational ethics) that direct attention to structurally different types of issues. Each disposition is defined apophatically (by what it refuses to do), equipped with a self-monitoring failure mode (hamartia), and orchestrated in sequence by role protocols. We evaluate this system on 50 merged pull requests across 7 repositories spanning 5 programming languages (Python, Go, C++, Java, Terraform), 5 organizations (2 enterprise, 3 open-source), and 2 temporal eras (pre-AI 2020, post-AI 2024--2026). The disposition system achieves 46% convergence with human reviewers (validating signal quality), identifies unique findings at a 75% rate, and produces no findings judged false-positive by the author across 601 total findings (inter-rater agreement was not assessed and remains a limitation). A controlled baseline comparison demonstrates that 51% of disposition findings are not produced by the same model using generic "expert reviewer" prompting, and these unique findings target structural, operational, and logical concerns rather than standard code-level issues. Preliminary cross-model validation (Claude Opus vs.\ GPT Codex 5.3-xhigh) on 3 PRs shows 100% framework-structure adherence with 39% finding-level agreement, suggesting the framework provides real behavioral constraint while preserving model-specific analytical perspective.

Large Language Models (LLMs) are increasingly integrated into software systems for diverse purposes, due to their versatility, flexibility, and ability to simulate human reasoning to some extent. However, poor integration of LLM inference in source code can undermine software system quality. Therefore, inadequate LLM integration coding practices must be documented to help developers mitigate such issues. Following our earlier work on LLM code smells, this paper consolidates and refines the concept by presenting a self-contained taxonomy and a catalog of nine LLM code smells. We also create SpecDetect4LLM, a static source code analysis tool for their detection, and conduct extensive empirical evaluations of its detection effectiveness (precision and recall) as well as the prevalence of LLM code smells across 692 open-source software projects (171,194 source files). Our results show that LLM code smells affect 73.5% of the analyzed systems, with a detection precision of 91.3% and a recall of 71.8%.

Linear properties are ubiquitous in the representations of language models; however, testing them experimentally remains a challenging task. This work focuses on relational linearity: the hypothesis that, for a fixed relation (e.g., "plays"), the unembedding of an object (e.g., "trumpet") can be predicted from the embedding of its subject (e.g.,"Miles Davis") by a linear map. We present an experimental method to test the formulation of relational linearity by Marconato et al. (2025). Specifically, we introduce a probing method, based on Kullback-Leibler divergence, to evaluate this property and examine its variation across layers and paraphrased relational queries. It is also more efficient than previous work; for example, it avoids the crude Jacobian approximations used in Linear Relational Embeddings by Hernandez et al. (2024). Our findings across four datasets show that relational linearity varies across models, exhibits layer-wise patterns consistent with prior observations about linguistic information in model representations, and is differently affected by changes in how the relation is phrased.

Third-party Python libraries introduce dependency management overhead, supply chain risk, and deployment friction in constrained environments. A natural question is how much of this ecosystem can be replicated using only Python's standard library -- and at what correctness and performance cost. We address this empirically through zerodep, a growing collection of single-file Python modules, each a stdlib-only reimplementation of a popular third-party library, developed with LLM assistance under strict constraints: no external imports, single file, drop-in API compatibility, and mandatory correctness validation against the reference library. Spanning over 40 modules across 12 categories -- including serialization, networking, cryptography, agent protocols, and text processing -- zerodep provides a controlled testbed for two interrelated questions: (1) Where does the stdlib suffice? and (2) Can LLMs effectively generate correct, performant code under tight symbolic constraints? Systematic benchmarking shows that stdlib-only implementations achieve performance parity (within 2x of the reference) in the majority of cases. The primary performance cliff is C-extension-backed computation (image processing, binary serialization, low-level crypto), not the inherent overhead of pure-Python third-party libraries. Conversely, many widely-used libraries carry architectural overhead that LLM-generated stdlib reimplementations avoid, yielding 5--115x speedups in several categories. We characterize the stdlib capability boundary across complexity tiers and library categories, discuss where LLM-assisted development succeeds and where it requires iterative human correction, and examine implications for dependency-free software engineering at scale. zerodep is open-source at https://github.com/Oaklight/zerodep.

Feedforward steering control is a key component of hierarchical control architectures for autonomous racing. The goal is to reduce steering corrections from the feedback controllers by predicting the vehicle's inverse lateral dynamics. This paper presents a systematic benchmark of two learning-based and two empirical (analytical) feedforward steering controllers. We introduce a new \acf{ehd} formulation based on a polynomial surface fit that captures velocity-dependent nonlinear steering behavior with minimal parametrization. We test the feedforward controllers in a high-fidelity simulation framework based on the real-world Abu Dhabi Autonomous Racing League competition, using a high-fidelity double-track vehicle dynamics simulator. Open-loop evaluation shows that the learning-based controllers achieve the lowest prediction errors; however, closed-loop testing reveals that this improved accuracy does not translate into superior path tracking performance or lap times, even after iterative fine-tuning. In contrast, the proposed EHD approach achieves the best overall closed-loop robustness and lap time, highlighting the necessity of evaluating feedforward strategies within the complete trajectory planning and control software stack. Our code is available at https://github.com/TUMRT/steering_ff_control.

Quadratic constraints (QCs) are widely used to characterize nonlinearities and uncertainties, but generic analytical characterizations can be conservative on bounded domains. This paper develops a framework for constructing verified quadratic characterizations of scalar relations in the two-dimensional real plane. Candidate quadratic inequalities are locally generated by solving convex quadratic programs using samples from the relation and exterior sample points. They are then verified globally using sum-of-squares certificates over an exact semialgebraic description or, in the case of nonpolynomial relations, over relaxed polynomial descriptions. The resulting verified constraints define a sound overapproximation of the scalar relations over the considered domains. These constraints are directly compatible with existing analysis frameworks based on QCs and pointwise integral quadratic constraints (IQCs) for static nonlinearities and uncertainties, and they can also be embedded in QC-based semidefinite programs for reachability and safety analysis of feedforward neural networks. For smooth activations such as $\tanh$, the method yields domain-dependent quadratic characterizations that constitute an alternative to generic sector- or slope-based descriptions. For ReLU networks, we give methods to reduce conservatism in QC-based reachability analysis of feedforward networks by exploiting dependencies between neurons and tighter local bounds. Numerical examples demonstrate improved reachability results for smooth activations, reduced conservatism for ReLU networks, and applicability beyond neural networks through an example involving saturation.

Modern deployments of Large Language Models (LLMs) increasingly require serving multiple models with diverse architectures, sizes, and specialization on shared, heterogeneous hardware. This setting introduces new challenges for resource allocation, dispatching, and scheduling, particularly under GPU memory constraints where partial CPU-GPU offloading and preemption become necessary. While existing systems primarily optimize throughput for a single model, comparatively little work addresses multi-model scheduling under these conditions. In this paper, we present an empirical study of how different LLMs behave across hardware platforms, focusing on the performance implications of layer offloading and preemption. We show that offloading leads to strongly non-linear and model-dependent degradation in decode throughput, with smaller models exhibiting sharper sensitivity to reduced GPU residency. We further demonstrate that preemption incurs substantial overhead, largely dominated by model state reload rather than key-value cache transfer, and that this cost varies significantly across models and hardware platforms. Additionally, we highlight the role of sequence length and interconnect bandwidth in amplifying data movement and execution inefficiencies. Based on these findings, we identify a set of key features that future schedulers must consider, including model-specific offloading sensitivity, workload characteristics, and the cost structure of preemption and data transfer. These insights provide guidance for the design of next-generation LLM serving systems capable of efficiently managing heterogeneous, multi-model workloads with hybrid CPU-GPU execution.

LLM-based chatbot agents increasingly process user requests by combining natural-language reasoning with external tools such as web browsing. These capabilities improve usability, but they also create attack surfaces when untrusted external content is processed as part of a user' s task. This paper studies a privacy-leakage attack chain based on indirect prompt injection in black-box chatbot environments, where the attacker has no access to model weights, system prompts, or agent implementation details including how a trajectory is actually managed during its processing for a query. We first analyze how an attacker can hijack an agent' s intended task by crafting external content that appears benign to the victim while inducing the agent to execute an attacker-defined objective. We then evaluate a new prompt-injection technique, called exemplification, which uses a bridge in the external content to reframe the user prompt and the benign beginning of the retrieved page as few-shot examples before appending the attacker' s objective. We compare its attack success rate with a prior fake-completion technique. Finally, we demonstrate a proof-of-concept data-exfiltration chain using fictitious personal information in a controlled setting. Our results suggest that prompt injection, jailbreak-style instruction steering, and web-tool invocation can be combined into a feasible privacy-leakage path in deployed chatbot agents.

Joint-Embedding Predictive Architectures (JEPA) are a promising framework for self-supervised video representation learning, yet the behavior of auxiliary objectives in small-scale Video-JEPA training is not well characterized. We report a small-scale empirical study of 18 auxiliary objective variants for Video-JEPA across two pretraining regimes: single-dataset (UCF-101) and mixed-dataset (UCF-101 + Something-Something V2 + ImageNet-100). We evaluate frozen representations on three complementary benchmarks: Diving-48 (fine-grained motion), SomethingSomething V2 (temporal reasoning), and ImageNet-100 (appearance). Our experiments suggest that many auxiliary objectives exhibit capacity trade-offs: gains on one downstream capability often coincide with degradation on another. We then study FWM-HW-LD (Factorized World-Model with Hard-Region-Weighted Latent Dynamics), a training-time objective that separates the latent representation into appearance and dynamics subspaces and applies hard-region weighting to both JEPA prediction errors and latent dynamics errors. In our mixed-dataset setting, FWM-HW-LD improves ImageNet-100 by +5.92 and SSv2 by +3.21 percentage points relative to the reference baseline, while remaining within 0.30 percentage points on Diving-48. These results indicate that latent factorization is a useful direction for studying auxiliary-objective trade-offs in Video-JEPA.

Recent advances in Time Series Foundation Models (TSFMs) promise zero-shot forecasting capabilities with minimal task-specific training. While these models have shown strong performance across generic benchmarks, their applicability in volatile, complex electricity markets remains underexplored. Addressing this gap, this study provides a systematic empirical evaluation of several TSFMs, specifically Chronos-2 and Chronos-Bolt (developed by Amazon), and TimesFM 2.5 (provided by Google), for forecasting Belgian day-ahead and imbalance electricity prices. For both considered markets, Chronos-2 in ARX mode produces the most accurate forecasts. Compared with the best ensemble prediction from other machine learning methods, Chronos-2's Mean Absolute Error (MAE) is 5% lower for the day-ahead market. In contrast, the model yields 10% higher MAE predicting imbalance prices across all forecast horizons, except for the two-hour-ahead horizon. Moreover, we find that TSFMs exhibit genuine zero-shot forecasting skills but still struggle under extreme market conditions.

Despite the remarkable success of transformer architectures in natural language processing, their scalability limitations remain poorly understood through systematic empirical analysis. This paper presents the first comprehensive large-scale evaluation of 118 transformer models across seven distinct architectural categories, revealing fundamental performance walls that manifest as hard deployment constraints. Our systematic benchmarking methodology uncovers a critical scalability crisis: while 88.1% of models successfully process sequences up to 512 tokens, this drops dramatically to 44.9% at 1024 tokens, with complete failure (0%) at 2048 tokens. Through rigorous analysis of loading times, memory consumption, and computational efficiency across sequence lengths from 128 to 2048 tokens, we demonstrate that compressed models achieve superior parameter efficiency (649.2 tokens/sec/M parameters) compared to large generative models (12.5 tokens/sec/M). Our findings challenge prevailing scaling assumptions and provide the first quantitative evidence that the theoretical O(n2) attention complexity translates into measurable performance walls. This work establishes new benchmarking methodologies for transformer evaluation and provides critical insights for practical deployment decisions in production environments.

Malformed, missing, or boundary-value inputs in microservice APIs can cascade across dependent services, threatening reliability. Robustness testing systematically exercises such inputs to expose server-side failures, but generating diverse, effective tests remains challenging. Large Language Models can generate such tests from API specifications; however, it is unknown whether different models and prompt strategies produce diverse failure sets or converge on the same failures. We report a controlled experiment applying 7 prompt strategies to 3 open-source LLMs (14B-70B parameters) targeting 2 architecturally distinct microservice systems: one Java monolingual (6 services, 9 failure modes) and one polyglot (27 services, 14 failure modes), yielding 38 valid runs and 663 generated tests. We find that prompt strategy explains more variation in diversity than model size: a Structured prompt collapses diversity entirely, while a single model varied across three prompt strategies achieves complete failure-mode coverage on one system, outperforming any multi-model ensemble under a fixed prompt. We introduce two strategies, Guided and GuidedFewShot, that embed a mutation taxonomy from prior robustness testing research as domain context. GuidedFewShot achieves the highest single-run coverage on both systems (5 of 9 and 8 of 14 failure modes) while maintaining low cross-model similarity. A key lesson is that taxonomy rules alone are insufficient: LLMs cannot distinguish key-absent from value-empty mutations without concrete examples. Findings replicate across both systems.

Validating training data for reasoning models typically requires expensive trial-and-error fine-tuning cycles. In this work, we investigate whether the utility of a reasoning dataset can be reliably predicted prior to training using intrinsic data metrics. We propose a suite of quantitative measures and evaluate their predictive power by fine-tuning 8B and 11B models on semantically distinct variants of a Polish reasoning dataset. Our analysis reveals that these intrinsic metrics demonstrate strong and significant correlations with downstream model performance. Crucially, we find that the predictors of utility are scale-dependent: smaller models rely on alignment-focused metrics to ensure precision, whereas larger models benefit from high redundancy, utilizing verbose traces to solve complex tasks. These findings establish a scale-aware framework for validating reasoning data, enabling practitioners to select effective training sets without the need for exhaustive empirical testing.

Automated detection of vulnerability-fixing commits (VFCs) is critical for timely security patch deployment, as advisory databases lag patch releases by a median of 25 days and many fixes never receive advisories. We present a comprehensive evaluation of code language model based VFC detection through a unified framework consolidating over 20 fragmented datasets spanning more than 180000 commits. Across over 180 experiments with fine-tuned models from 125 M to 14 B parameters, we find no evidence that models acquire transferable security-relevant code understanding from code changes alone. When commit messages are available, they dominate model attention, and when removed, an attribution analysis shows that enriching diffs with additional intra-procedural semantic context does not shift model attention toward the code changes. Group-stratified evaluation exposes approximately 17% performance drops compared to random splits, while temporal splits on aggregated datasets prove unreliable due to compositional shift in the underlying project distributions. At a false positive rate of 0.5% all fine-tuned code-only models miss over 93% of vulnerabilities. Larger and more diverse training data or generative approaches show preliminary improvements but do not resolve the underlying limitations. To support future research on code-centric VFC detection, we release our unified framework and evaluation suite.

The literature on how large language models handle conflict between their training knowledge and a contradicting document presents a persistent empirical contradiction: some studies find models stubbornly retain their trained answers, ignoring provided documents nearly half the time, while others find models readily defer to the document, following context approximately 96% of the time. We argue these contradictions dissolve once one recognises that prior experiments have studied three qualitatively distinct processing situations without distinguishing them. We propose a three-regime framework: Regime 1 (single-source updating, dominant predictor: evidence coherence), Regime 2 (competitive integration, dominant predictor: parametric certainty), and Regime 3 (task-appropriate selection, dominant predictor: task knowledge requirement). We formalise a distinction between parametric strength (exposure frequency) and parametric uniqueness (encoding consistency), showing empirically that these are orthogonal dimensions (r = -0.002, p = .97) with strength as the operative predictor in stable factual domains. We validate the framework across Claude Sonnet 4.6, GPT-5.5, Gemini 2.5 Flash, Llama 4 Maverick, and DeepSeek V3 using 9,970 API calls in three experimental phases. GEE logistic regression confirms the predicted Regime 2 certainty gradient for all five models (beta = -0.38 to -0.50, all p <= .013, BH-FDR corrected). A Regime 3 ablation shows task framing alone flips context-following from near-100% (contextual knowledge condition) to 6-71% (parametric knowledge condition), with all five models significant (p < .001). The certainty gradient is robust to multinomial outcome modeling, sensitivity analyses for hedging responses, and FDR correction.

Agent evaluation requires assessing complex multi-step behaviors involving tool use and intermediate reasoning, making it costly and expertise-intensive. A natural question arises: can frontier coding assistants reliably automate this evaluation process? Our study shows that simply prompting coding assistants is insufficient for this task. Without domain-specific evaluation knowledge, frontier coding assistants achieve only a 30% execution success rate and produce over-engineered evaluations averaging 12+ metrics per agent, indicating that strong coding ability does not automatically translate to reliable agent evaluation. We introduce EvalAgent, an AI assistant that automates the end-to-end agent evaluation pipeline. EvalAgent encodes evaluation domain expertise as evaluation skills (procedural instructions, reusable code and templates, and dynamically retrieved API documentation) that compose into a trace-based pipeline producing complete evaluation artifacts including metrics, executable code, and reports. To systematically assess generated evaluations, we introduce a meta-evaluation framework alongside AgentEvalBench, a benchmark comprising 20 agents, each paired with evaluation requirements and test scenarios. We further propose the Eval@1 metric to measure whether generated evaluation code both executes and yields meaningful results on the first run. Our experiments show that EvalAgent produces focused evaluations, improving Eval@1 from 17.5% to 65%, and achieving 79.5% human expert preference over baseline approaches. Further ablation studies show that evaluation skills are critical for handling complex evaluation: removing them causes Eval@1 to drop significantly from 65% to 30%.

This work proposes Attractor-Vascular Coupling Theory (AVCT), a mathematical framework showing that cardiac attractor geometry encodes blood pressure (BP) information sufficient for AAMI-standard estimation, and validates the theory through a calibrated cuffless BP model using photoplethysmography (PPG). AVCT is grounded in Cardiac Stability Theory and operationalized using Takens delay embedding and attractor morphology extraction. Two theorems, one proposition, and one corollary formally justify the use of PPG attractor features for BP estimation and predict the feature-importance hierarchy. A LightGBM model trained on pulse transit time (PTT) and Cardiac Stability Index (CSI) attractor features under single-point calibration was evaluated using strict leave-one-subject-out cross-validation (LOSO-CV) on 46 subjects from BIDMC ICU (n = 9) and VitalDB surgical data (n = 37), comprising 29,684 windows. The model achieved systolic BP (SBP) mean absolute error (MAE) of 2.05 mmHg and diastolic BP (DBP) MAE of 1.67 mmHg, with correlations r = 0.990 and r = 0.991, satisfying the AAMI/IEEE SP10 requirement of MAE below 5 mmHg. Median per-subject MAE was 1.87/1.54 mmHg, and 70%/76% of subjects individually satisfied AAMI criteria. A PPG-only ablation using nine smartphone attractor features matched the ECG+PPG model within 0.05 mmHg, demonstrating that clinical-grade BP tracking is achievable using only a smartphone camera while surpassing prior generalized LOSO-CV results using fewer sensors. All four AVCT predictions were quantitatively confirmed, with 91.5% error reduction from uncalibrated to calibrated estimation (epsilon_cal = 0.915). Unlike post-hoc explainable AI methods, AVCT predicts features satisfying the architectural faithfulness criterion of the Explainable-AI Trustworthiness (EAT) framework and grounding BP estimation in nonlinear dynamical systems theory.

Large Language Models(LLMs) are increasingly explored for cybersecurity applications such as vulnerability detection. In the domain of threat modelling, prior work has primarily evaluated a number of general-purpose Large Language Models under limited prompting settings. In this study, we extend the research area of structured threat modelling by systematically evaluating domain-adapted language models of different sizes to their general counterparts. We use both LLMs and Small Language Models(SLMs) that were domain adapted to telecommunications and cybersecuirty. For the structured threat modelling, we selected the widely used STRIDE approach and the application area is 5G security. We present a comprehensive empirical evaluation using 52 different configurations (on 8 different language models) to analyze the impact of 1) domain adaptation, 2) model scale, 3) decoding strategies (greedy vs. stochastic sampling), and 4) prompting technique on STRIDE threat classification. Our results show that domain-adapted models do not consistently outperform their general-purpose counterparts, and decoding strategies significantly affect model behavior and output validity. They also show that while larger models generally achieve higher performance, these gains are neither consistent nor sufficient for reliable threat modelling. These findings highlight fundamental limitations of current LLMs for structured threat modelling tasks and suggest that improvements require more than additional training data or model scaling, motivating the need for incorporating more task-specific reasoning and stronger grounding in security concepts. We present insights on invalid outputs encountered and present suggestions for prompting tailored specifically for STRIDE threat modelling.

Qualitative analysis plays a pivotal role in understanding the human and social aspects of software engineering. However, it remains a demanding process shaped by the subjective interpretation of individual researchers and sensitive to methodological choices such as prompt design. Recent advancements in Large Language Models (LLMs) offer promising opportunities to support this type of analysis, although their reliability in reproducing human qualitative reasoning under varying prompting conditions remains largely untested. This study presents a controlled empirical evaluation of three LLMs -- Claude Haiku, DeepSeek-Chat, and Gemini 2.5 Flash -- across two prompt engineering strategies (zero-shot and multi-shot closed coding), using Cohen's kappa as the primary agreement metric over ten independent runs per configuration. Results suggest that multi-shot prompting significantly improves agreement for Claude Haiku (Delta kappa = +0.034, Wilcoxon p = 0.004) but not for DeepSeek-Chat or Gemini 2.5 Flash. Intra-model stability varies substantially -- DeepSeek-Chat and Claude Haiku exhibit the lowest variance (SD approx. 0.017), while Gemini 2.5 Flash is the least stable (SD = 0.038). A systematic over-prediction of "Sharing Negative Feedback" is identified across all models (bias ratios up to 5.25x), alongside consistent under-prediction of "Expressing Concerns." Collectively, these findings provide empirical evidence for prompt engineering guidelines in LLM-assisted qualitative coding for software engineering research.

Adaptive rank allocation for LoRA, allocating more parameters to important layers and fewer to unimportant ones, consistently improves efficiency under supervised fine-tuning (SFT). We investigate whether this success transfers to reinforcement learning, specifically Group Relative Policy Optimization (GRPO). Using gradient-magnitude profiling on Qwen 2.5 1.5B with GSM8K, we find that it does not: proportional rank allocation degrades accuracy by 4.5 points compared to uniform allocation (70.0% vs. 74.5%), despite using identical parameter budgets. We identify two mechanisms behind this failure. First, the gradient landscape under GRPO is fundamentally flatter than under SFT, the max-to-min layer importance ratio is only 2.17x, compared to >10x reported in SFT literature. All layers carry meaningful gradient signal; none are truly idle. Second, we discover a gradient amplification effect: non-uniform allocation widens the importance spread from 2.17x to 3.00x, creating a positive feedback loop where high-rank layers absorb more gradient while low-rank layers are progressively silenced. Our results suggest that gradient importance does not predict capacity requirements under RL, and that naive transfer of SFT-era rank allocation to alignment training should be avoided.

Large Language Models (LLMs) have achieved great improvements in recent years. Nevertheless, it still remains unclear how good LLMs are for reasoning tasks, especially for long-chain ones. In this paper, we evaluate LLMs' performance on the simplest yet long-chain reasoning task, namely the Equivalence Class Problem (ECP), i.e., determining whether two variables are equal given a set of randomly generated equivalence relations. We consider both reasoning and non-reasoning representative LLMs over a large variety of problem instances, ranging over different numbers of variables, connectivity probabilities, prompts, and other factors. The experimental results show that non-reasoning LLMs fail ECP, while reasoning models are significantly better but still struggle to completely solve this problem. Interestingly, considering various connectivity probabilities with a fixed number of variables, we observe that, for non-reasoning models, the hardest problem instances coincide with the phase transition point of ln n/(n-1), suggesting the chaos of the problem; in contrast, for reasoning models, the hardest ones coincide with the biggest diameter, suggesting the reasoning difficulty of the problem.

Large language models (LLMs) exhibit a wide range of human-like behaviors, from expressing thoughts and emotions, to engaging in relationship-building with users, to refusing requests and maintaining boundaries. Despite their prevalence, researchers and practitioners lack methods and empirical insights to make informed decisions about when and what types of human-like behaviors LLMs should exhibit. To fill this gap, we present a multi-dimensional analysis of the prevalence, potential effects, and controllability of these behaviors using LLM-as-a-judge and human evaluation. Across 21,000 multi-turn conversations from four widely used models (gpt-4o, gpt-4.1-mini, claude-sonnet-4.6, gemini-2.5-flash), we find that human-like behaviors are pervasive but vary across models and user factors (conversation goals and user profiles). In terms of perceived appropriateness, human evaluators judged self-referential and relationship-building behaviors as less appropriate from LLMs than from humans, but boundary-maintaining behaviors more appropriate from LLMs than from humans. Finally, we show that system prompting can control these behaviors, though it requires careful evaluation to avoid unintended effects. We discuss the implications of our findings and provide recommendations for responsible LLM design and evaluation.

Understanding the topology of decision regions is central to explaining the inner workings of deep neural networks. Prior empirical work has provided evidence that these regions are path connected. We study a stronger topological question: whether closed loops inside a decision region can be contracted without leaving that region. To this end, we propose an iterative quad-mesh filling procedure that constructs a finite-resolution label-preserving surface bounded by a given loop and lying entirely within the same decision region. We further connect this construction to natural Coons patches in order to quantify its deviation from a canonical geometric interpolation of the loop. By evaluating our method across several modern image-classification models, we provide empirical evidence supporting the hypothesis that decision regions in deep neural networks are not only path connected, but also simply connected.

Synthetic data has transformed language model training, yet its role in time series forecasting remains poorly understood. We present a large-scale empirical study: nine experiment groups, 4,218 runs systematically evaluating synthetic time series augmentation across five architectures, four synthetic signals and seven datasets. The effect is sharply architecture-conditional: channel-mixing models (TimesNet, iTransformer) benefit in the majority of trials, while channel-independent models (DLinear, PatchTST) are consistently degraded. In selected low-resource settings the gains are striking: TimesNet trained on only 10% of Weather data with synthetic augmentation surpasses the full-data baseline (4 of 16 sparsity-dataset combinations). Averaged across all architectures, augmentation hurts in 67% of trials. We further find that only the Seasonal-Trend generator reliably helps across the tested benchmarks, and that hard curriculum switching is actively harmful (+24% MSE degradation). These results provide concrete, actionable guidelines on how to use synthetic data: use synthetic augmentation with channel-mixing architectures, use gradual annealing schedules, and treat low-resource augmentation as architecture- and dataset-dependent. Code is available at \href{https://github.com/hugoiscracked/synthetic-ts/tree/main}

Large language model (LLM)-based coding assistants have made substantial progress, yet most systems remain reactive, requiring developers to explicitly formulate their needs. Proactive coding assistants aim to infer latent developer intent from integrated development environment (IDE) interactions and repository context, thereby reducing interaction overhead and supporting more seamless assistance. However, research in this direction is limited by the scarcity of large-scale real-world developer behavior data. Existing studies therefore often rely on LLM-simulated IDE traces, whose fidelity to real development behavior remains unclear. In this paper, we investigate this simulation-to-reality gap through a large-scale empirical study. We collect real IDE interaction traces from 1{,}246 experienced industry developers over three consecutive days using a custom Visual Studio Code extension, and construct paired LLM-simulated traces for controlled comparison. Our analysis shows that simulated traces differ substantially from real traces in behavioral diversity, temporal structure, and exploratory patterns. Based on the collected data, we introduce \textbf{ProCodeBench}, a real-world benchmark for proactive intent prediction. Experiments with representative LLMs, retrieval-augmented methods, and agentic baselines show that current approaches remain far from reliable under real IDE traces, suggesting that simulation-based evaluation can overestimate real-world performance. Finally, our training study shows that simulated data cannot replace real data, but can complement it when used before real-world fine-tuning. These findings highlight the importance of real developer behavior data for evaluating and training proactive coding assistants.

As Vision-Language Models (VLMs) are increasingly deployed as autonomous cognitive cores for embodied assistants, evaluating their privacy awareness in physical environments becomes critical. Unlike digital chatbots, these agents operate in intimate spaces, such as homes and hospitals, where they possess the physical agency to observe and manipulate privacy-sensitive information and artifacts. However, current benchmarks remain limited to unimodal, text-based representations that cannot capture the demands of real-world settings. To bridge this gap, we present ImmersedPrivacy, an interactive audio-visual evaluation framework that simulates realistic physical environments using a Unity-based simulator. ImmersedPrivacy evaluates physically grounded privacy awareness across three progressive tiers that test a model's ability to identify sensitive items in cluttered scenes, adapt to shifting social contexts, and resolve conflicts between explicit commands and inferred privacy constraints. Our evaluation of 12 state-of-the-art models reveals consistent deficits. In cluttered scenes, all models exhibit monotonic performance decay as scene complexity grows due to perceptual deficit. When social context shifts, no model exceed 65% selection accuracy. Under conflicting commands, the best model gemini-3.1-pro perfectly balances task completion and privacy preservation in only 51% of cases. These findings reveal that current VLMs in the physical world suffer from perceptual fragility and fail to let their knowledge of privacy cues govern their situated behavior. Our code and data is available at https://github.com/immersed-privacy/immersed-privacy .

This revision updates a pop-to-jazz chord-generation rehearsal study. Best-epoch metrics still show that modest pop rehearsal preserves pop accuracy while improving jazz prediction, but v2 corrects released-checkpoint selection: the released F1 equals Phase 0, F2 had a transcription error, and ft-pop80-v2 restores a hash-distinct jazz-adapted F1 across 3 seeds.

As LLMs enter conflict monitoring, understanding systematic distortions in their outputs is critical for humanitarian accountability. We evaluate four vanilla open-weight models Gemma 3 4B, Llama 3.2 3B, Mistral 7B, and OLMo 2 7B and two domain-adapted models, AfroConfliBERT and AfroConfliLLAMA, on Nigeria and Cameroon conflict-event classification against ACLED, a gold-standard dataset with multi-stage verification. We find a bifurcated divergence in normative directionality. Open-weight models exhibit statistically significant False Illegitimation bias: Gemma misclassifies to 18.29% of legitimate battles as civilian-targeted violence while making zero False Legitimation errors. By contrast, AfroConfliBERT and AfroConfliLLAMA achieve near-directional neutrality, with Legitimization Bias differences indistinguishable from zero. Yet domain adaptation does not eliminate actor-based selection bias. Both adapted models show statistically significant actor bias comparable to vanilla LLMs; in Nigeria, state actors are legitimized 36.5% more often than non-state actors in identical tactical contexts. Open-weight outputs are also fragile to geography-specific lexical framing: delegitimizing phrases produce flip rates up to 66.7% in Cameroon and 34.2% in Nigeria, while perturbations salient in one context may not matter in another. Error trace profiling shows models mask normative bias through unfaithful rationale confabulations. In contrast, AfroConfliBERT and AfroConfliLLAMA are largely robust, with near-zero flip rates across perturbation categories. Overall, current models are not ready for unsupervised deployment in conflict monitoring. We call for fairness-aware fine-tuning to reduce actor-based selection bias, mandatory adversarial robustness evaluation against lexical manipulation, and context-specific human-in-the-loop oversight calibrated to regional difficulty.

Neural operators have emerged as a powerful tool for learning mappings between infinite-dimensional function spaces. However, their approximation properties in Sobolev norms remain poorly quantified, even though these norms control both function values and derivatives and are the natural metrics for PDE well-posedness, stability, and generalization. We develop a functional-analytic framework for operator learning in Sobolev spaces and connect it to the numerical behavior of Fourier Neural Operators (FNOs) on a prototypical PDE. First, for a continuous nonlinear operator $\mathcal{G}: H^{s}(D)\to H^{t}(D')$ with $s > d/2$ and inputs restricted to a compact subset of $H^{s}(D)$, we prove that $\mathcal{G}$ can be uniformly approximated in $H^{t}$-norm by a neural operator with $\mathcal{O}(\varepsilon^{-d/s})$ trainable parameters. This yields an explicit complexity--error relation of the form $\|\mathcal{G}-\mathcal{G}_θ\|_{H^{t}} \lesssim C N^{-s/d}$. We then study the one-dimensional viscous Burgers solution operator $\mathcal{G}: u_{0}\mapsto u(\cdot,1)$ on a bounded $H^{1}$-ball and train FNOs with an $H^{1}$-loss. Across a sweep of model sizes, we obtain test $H^{1}$-errors down to $\mathcal{O}(10^{-7})$ and relative errors of order $10^{-3}$, with predictions accurately matching both solutions and spatial derivatives on held-out data. A log-log plot of Sobolev error versus parameter count exhibits an approximate power law $\|\mathcal{G}-\mathcal{G}_θ\|_{H^{1}} \approx C N^{-α}$ with empirical exponent $α\approx 1.4$, and long-horizon training reveals optimization instabilities in large FNOs, providing quantitative evidence that Sobolev-space approximation theory meaningfully predicts neural-operator scaling behavior.

Causal Discovery (CD) is a powerful framework for scientific inquiry. Yet, its practical adoption is hindered by a reliance on strong, often unverifiable assumptions and a lack of robust performance assessment. To address these limitations and advance empirical CD evaluation, we present TCD-Arena, a modularized, highly customizable, and extendable testing kit to assess the robustness of time series CD algorithms against stepwise more severe assumption violations. For demonstration, we conduct an extensive empirical study comprising around 30 million individual CD attempts and reveal nuanced robustness profiles for 33 distinct assumption violations. Further, we investigate CD ensembles and find that they have the potential to improve general robustness, which has implications for real-world applications. With this, we strive to ultimately facilitate the development of CD methods that are reliable for a diverse range of synthetic and potentially real-world data conditions.

Healthcare automation is shaped by local procedures and organizational constraints, so agent capabilities rarely transfer unchanged across settings. Agent skills, self-contained directories that package reusable procedures for AI agents, are emerging as a procedural layer for adapting healthcare agents across diverse healthcare settings. We present the first empirical analysis of healthcare agent skills, drawing on 557 healthcare-related skills filtered from 58,159 public skills on ClawHub and annotated along ten dimensions covering function, deployment context, autonomy, and safety. We find that public healthcare skills emphasize patient-facing workflow automation and monitoring rather than the diagnostic and treatment-oriented tasks foregrounded in healthcare-agent research; coverage of the healthcare lifecycle and specialized clinical inputs remains uneven; and general technical risk does not reliably capture clinical risk. These findings position healthcare skills as a procedural layer not yet addressed by current benchmarks and risk frameworks.

Large language models (LLMs) have shown promise as interactive agents that solve tasks through extended sequences of environment interactions. While prior work has primarily focused on system-level optimizations or algorithmic improvements, the role of task horizon length in shaping training dynamics remains poorly understood. In this work, we present a systematic empirical study that examines horizon length through controlled task constructions. Specifically, we construct controlled tasks in which agents face identical decision rules and reasoning structures, but differ only in the length of action sequences required for successful completion. Our results reveal that increasing horizon length alone constitutes a training bottleneck, inducing severe training instability driven by exploration difficulties and credit assignment challenges. We demonstrate that horizon reduction is a key principle to address this limitation, stabilizing training and achieving better performance in long-horizon tasks. Moreover, we find that horizon reduction is related to stronger generalization across horizon lengths: models trained under reduced horizons generalize more effectively to longer-horizon variants at inference time, a phenomenon we refer to as horizon generalization.

Recent advancements in Large Language Models (LLMs) have demonstrated significant potential across a wide range of software engineering tasks, including software design, an area traditionally regarded as highly dependent on human expertise and judgment. However, there has been little research focusing on how LLMs are used in software design, nor on the associated benefits and drawbacks. This paper aims to bridge this gap by empirically investigating how software developers utilize LLMs in the context of software design. We conduct a mixed-methods study, combining a mining study of 291 developer-ChatGPT conversations shared on GitHub with a survey of 65 software practitioners. Our findings reveal nine distinct categories of design tasks supported by ChatGPT, including architecture design, data model design, and the use of design patterns. We further characterize developer-ChatGPT interactions, showing that developers primarily use ChatGPT for knowledge acquisition and design-related code generation, with most tasks situated at the detailed design level. The study identifies seven key benefits of utilizing LLMs in software design as perceived by developers, such as better technology selection and the early detection of design flaws. We also uncover six limitations, including the generation of overly lengthy and difficult-to-read outputs, the creation of inexecutable or incorrect code, and a heavy reliance on context that can lead to hallucinated results. These findings provide an evidence-based characterization of current LLM use in software design from both open-source and practitioner perspectives, highlighting a tension between perceived benefits and limitations, which lays a foundation for future research and the development of effective techniques and tools to integrate LLMs into software design practices.

Structured information extraction from long, multilingual scanned financial documents is a core requirement in industrial KYC and compliance workflows. These documents are typically non machine readable, noisy, and visually heterogeneous. They usually span dozens of pages while containing only sparse task relevant information. Although recent vision-language models achieve strong benchmark performance, directly applying them end to end to full financial reports often leads to unreliable extraction under real world conditions. We present a multistage extraction framework that integrates image preprocessing, multilingual OCR, hybrid page-level retrieval, and compact VLM-based structured extraction. The design separates page localization from multimodal reasoning, enabling more accurate extraction from complex multipage documents. We evaluated the framework on 120 production KYC documents comprising about 3000 multilingual scanned pages. Across multiple OCR-VLM combinations, the proposed pipeline consistently outperforms direct PDF-to-VLM baselines, improving field-level accuracy by up to 31.9 percentage points. The best configuration, PaddleOCR with MiniCPM2.6, achieves 87.27 percent accuracy. Ablation studies show that page-level retrieval is the dominant factor in performance improvements, particularly for complex financial statements and non-English documents.

Improving the Theory of Mind (ToM) capability of Large Language Models (LLMs) is crucial for effective social interactions between these AI models and humans. However, the existing benchmarks often measure ToM capability improvement through story-reading, multiple-choice questions from a third-person perspective, while ignoring the first-person, dynamic, and open-ended nature of human-AI (HAI) interactions. To directly examine how ToM improvement techniques benefit HAI interactions, we first proposed the new paradigm of interactive ToM evaluation with both perspective and metric shifts. Next, following the paradigm, we conducted a systematic study of four representative ToM enhancement techniques using both four real-world datasets and a user study, covering both goal-oriented tasks (e.g., coding, math) and experience-oriented tasks (e.g., counseling). Our findings reveal that improvements on static benchmarks do not always translate to better performance in dynamic HAI interactions. This paper offers critical insights into ToM evaluation, showing the necessity of interaction-based assessments in developing next-generation, socially aware LLMs for HAI symbiosis.

Despite AI tools becoming increasingly embedded in academic practice, little is known about how university students integrate them into their writing processes. We examine how students engage with AI across different writing tasks, and how this engagement is shaped by individual factors including AI literacy, writing confidence, trust, authorship concerns, and motivation. Study1 surveys 107 UK university students to map task-specific and co-occurring patterns of AI use across five writing stages (ideation, sourcing, planning, drafting, and reviewing) and their associations with individual factors. Study2 complements this by exploring how these patterns can be assembled in practice, through interviews with 12 postgraduates reflecting on their established use of AI in assessed writing. Together, the studies suggest that AI integration is selective and heterogeneous, forming three recurring and value-oriented configurations: (1) early-stage (learning-oriented), where tools support exploration and understanding; (2) late-stage (quality-oriented), where tools support drafting and refinement; and (3) peripheral (productivity-oriented), where tools are used to reduce friction and sustain momentum across the process. We offer a workflow-level account of AI-supported academic writing, showing how students navigate competing priorities of learning, quality, productivity, and authorship, and how they evaluate and take responsibility for AI-generated outputs.

Tian (2025) proves a repulsion theorem (Theorem 6) for the matrix $B = (\widetilde{F}^\top \widetilde{F} + ηI)^{-1}$ during the interactive feature-learning stage of grokking: similar features have negative off-diagonal entries $B_{j\ell}$, producing an effective repulsive force that drives them apart. However, the theorem does not specify when this mechanism becomes empirically observable, nor whether it leaves a measurable spectral signature in the parameter updates. We test this directly on Tian's modular addition setup ($M = 71$, $K = 2048$, MSE loss) and observe a clear structure-mechanism dissociation. The predicted sign rule holds robustly on the top-200 most-similar feature pairs across activations (empirical sign-match rising from 0.865 to 0.985 on $σ= x^2$ across 5 seeds, and saturating at 1.000 on $σ= \operatorname{ReLU}$). However, the spectral signature in the parameter updates is strongly activation-dependent. With $σ= x^2$, a simple slope detector on the rolling eigengap $σ_2 / σ_3$ of $ΔW$ fires in 15/15 grokking seeds at epoch 174 (IQR [173,174]) and in 0/15 non-grokking controls, with 229$\times$ late-stage magnitude separation; the spectrum is rank-2. In contrast, with $σ= \operatorname{ReLU}$, the detector never fires and the spectrum remains effectively rank-1. This dissociation aligns with Tian's Theorem 5 distinction between focused (power-law) and spreading (ReLU) memorization: while the sign structure of $B$ depends only on $\widetilde{F}^\top \widetilde{F}$, how feature repulsion translates into weight updates critically depends on the activation derivative $σ'$.

Artificial Intelligence and Machine Learning (AI/ML) models used in clinical settings are increasingly deployed to support clinical decision-making. However, when training data become stale due to changes in demographics, environment, or patient behaviors, model performance can degrade substantially. While updating models with new training data is necessary, such updates may also introduce new risks. We evaluated the proposed monitoring framework on four publicly available U.S.-based Type 1 Diabetes datasets containing high-resolution continuous glucose monitoring (CGM) data, comprising approximately 11,300 weekly observations from 496 participants under 20 years of age. All datasets included structured sociodemographic information. Using the prediction of severe hyperglycemia events in children with type 1 diabetes as a case study, we examine how different model update strategies can adversely affect model stability (e.g., by causing predictions to "flip" for a large number of cases after an update), increase arbitrariness in predictions, or worsen accuracy equity and the balance of error rates across subpopulations. We propose multiple dimensions for continuous monitoring to detect these issues and argue that such monitoring is essential for the development of trustworthy clinical decision support systems.

Large Language Models (LLMs) are increasingly deployed in persona-driven applications such as education, customer service, and social platforms, where models are prompted to adopt specific personas when interacting with users. While persona conditioning can improve user experience and engagement, it also raises concerns about how personality cues may interact with gender biases and stereotypes. In this work, we present a controlled study of persona-conditioned story generation in English and Hindi, where each story portrays a working professional in India producing context-specific artifacts (e.g., lesson plans, reports, letters) under systematically varied persona gender, occupational role, and personality traits from the HEXACO and Dark Triad frameworks. Across 23,400 generated stories from six state-of-the-art LLMs, we find that personality traits are significantly associated with both the magnitude and direction of gender bias. In particular, Dark Triad personality traits are consistently associated with higher gender-stereotypical representations compared to socially desirable HEXACO traits, though these associations vary across models and languages. Our findings demonstrate that gender bias in LLMs is not static but context-dependent. This suggests that persona-conditioned systems used in real-world applications may introduce uneven representational harms, reinforcing gender stereotypes in generated educational, professional, or social content.

Graph-based anti-money laundering (AML) systems on blockchain networks can score suspicious activity at two granularity levels -- transactions or actor addresses -- yet compliance action is conducted per actor. This paper contributes an evaluation methodology for measuring how scoring granularity affects investigation queue composition under fixed review budgets. We formalize the evaluation through a projection framework mapping transaction-level scores to the actor-level action unit via four aggregation operators, and introduce budgeted investigation metrics -- yield@budget, burden decomposition, and case fragmentation. Using the public Elliptic++ Bitcoin dataset (203,769 transactions; 822,942 address occurrences), we train independent random forest classifiers at each level under a causal temporal protocol and compare review queues through Jaccard overlap, burden decomposition, and feature-matching ablations. At one-percent budget, temporal evaluation yields mean Jaccard of 0.374 (SD 0.171); static pooled evaluation yields 0.087 (95% CI [0.079, 0.094]). An enriched address model receiving all 237 features produces even lower overlap (Jaccard=0.051), with 4.3% illicit per 100 reviews versus 30.2% for the transaction-projected queue. Address-level detection value is temporally concentrated: two timesteps exceed 91% illicit per 100 reviews while the static burden is only 3.4%. A fixed hybrid policy underperforms the best single-level queue by 5.05pp (CI [-10.2pp, -0.9pp]). These findings establish that scoring granularity is a consequential design variable for AML investigation systems -- same data, same budget, different queues, different addresses investigated.

Large language models (LLMs) have demonstrated strong performance on a wide range of software engineering tasks, including code generation and analysis. However, most prior work relies on cloud-based models or specialized hardware, limiting practical applicability in privacy-sensitive or resource-constrained environments. In this paper, we present a systematic empirical evaluation of two locally deployed LLMs, LLaMA 3.2 and Mistral, for real-world Python bug detection using the BugsInPy benchmark. We evaluate 349 bugs across 17 projects using a zero-shot prompting approach at the function level and an automated keyword-based evaluation framework. Our results show that locally executed models achieve accuracy between 43% and 45%, while producing a large proportion of partially correct responses that identify problematic code regions without pinpointing the exact fix. Performance varies significantly across projects, highlighting the importance of codebase characteristics. The results demonstrate that local models can identify a meaningful share of bugs, though precise localization remains difficult for locally executed LLMs, particularly when handling complex and context dependent bugs in realistic development scenarios.

Geographic context is often consider relevant to motor insurance risk, yet public actuarial datasets provide limited location identifiers, constraining how this information can be incorporated and evaluated in claim-frequency models. This study examines how geographic information from alternative data sources can be incorporated into actuarial models for Motor Third Party Liability (MTPL) claim prediction under such constraints. Using the BeMTPL97 dataset, we adopt a zone-level modeling framework and evaluate predictive performance on unseen postcodes. Geographic information is introduced through two channels: environmental indicators from OpenStreetMap and CORINE Land Cover, and orthoimagery released by the Belgian National Geographic Institute for academic use. We evaluate the predictive contribution of coordinates, environmental features, and image embeddings across three baseline models: generalized linear models (GLMs), regularized GLMs, and gradient-boosted trees, while raw imagery is modeled using convolutional neural networks. Our results show that augmenting actuarial variables with constructed geographic information improves accuracy. Across experiments, both linear and tree-based models benefit most from combining coordinates with environmental features extracted at 5 km scale, while smaller neighborhoods also improve baseline specifications. Generally, image embeddings do not improve performance when environmental features are available; however, when such features are absent, pretrained vision-transformer embeddings enhance accuracy and stability for regularized GLMs. Our results show that the predictive value of geographic information in zone-level MTPL frequency models depends less on model complexity than on how geography is represented, and illustrate that geographic context can be incorporated despite limited individual-level spatial information.

Theory and empirical research on management teams' influence on firm performance have witnessed continuous development, and by now incorporate numerous details. Classic, experiment-based studies examining social systems collect vast amount of data, but often times investigate only the first one or two modes of the distribution of measured variables, and experience difficulty in analyzing the effect of context. For example, in functional diversity research, management teams are described by measures incorporating complex distributions of capabilities of individual managers and teams of managers. To investigate the effect of hidden distributions, and the effect of functional diversity composition on team communication and performance, we developed an agent-based model, and conducted a series of simulation experiments. Modeling results show that depending on the context, such as communication scheme among interacting agents, or their functional composition, intrapersonal functional diversity (IFD), and dominant function diversity (DFD) might enhance or reduce performance and communication among agents. Furthermore, simulation results also suggest that a third measure is required alongside IFD and DFD capturing the aggregate expertise of the team to comprehensively account for empirical findings.

Synthetic data generation offers promise for addressing data scarcity and privacy concerns in educational technology, yet practitioners lack empirical guidance for selecting between traditional resampling techniques and modern deep learning approaches. This study presents the first systematic benchmark comparing these paradigms using a 10,000-record student performance dataset. We evaluate three resampling methods (SMOTE, Bootstrap, Random Oversampling) against three deep learning models (Autoencoder, Variational Autoencoder, Copula-GAN) across multiple dimensions: distributional fidelity (Kolmogorov-Smirnov distance, Jensen-Shannon divergence), machine learning utility such as Train-on-Synthetic-Test-on-Real scores (TSTR), and privacy preservation (Distance to Closest Record). Our findings reveal a fundamental trade-off: resampling methods achieve near-perfect utility (TSTR: 0.997) but completely fail privacy protection (DCR ~ 0.00), while deep learning models provide strong privacy guarantees (DCR ~ 1.00) at significant utility cost. Variational Autoencoders emerge as the optimal compromise, maintaining 83.3% predictive performance while ensuring complete privacy protection. We also provide actionable recommendations: use traditional resampling for internal development where privacy is controlled, and VAEs for external data sharing where privacy is paramount. This work establishes a foundational benchmark and practical decision framework for synthetic data generation in learning analytics.

Downhole drilling telemetry presents a fundamental labeling asymmetry: surface sensor data are generated continuously at 1Hz, while labeled downhole measurements are costly, intermittent, and scarce. Current machine learning approaches for downhole metric prediction universally adopt fully supervised training from scratch, which is poorly suited to this data regime. We present the first empirical evaluation of masked autoencoder (MAE) pretraining for downhole drilling metric prediction. Using two publicly available Utah FORGE geothermal wells comprising approximately 3.5 million timesteps of multivariate drilling telemetry, we conduct a systematic full-factorial design space search across 72 MAE configurations and compare them against supervised LSTM and GRU baselines on the task of predicting Total Mud Volume. Results show that the best MAE configuration reduces test mean absolute error by 19.8% relative to the supervised GRU baseline, while trailing the supervised LSTM baseline by 6.4%. Analysis of design dimensions reveals that latent space width is the dominant architectural choice (Pearson $r = -0.59$ with test MAE), while masking ratio has negligible effect, an unexpected finding attributed to high temporal redundancy in 1Hz drilling data. These results establish MAE pretraining as a viable paradigm for drilling analytics and identify the conditions under which it is most beneficial.

The rapid growth of chemical literature has generated vast amounts of unstructured data, where reaction information is particularly valuable for applications such as reaction predictions and drug design. However, the prohibitive cost of expert annotation has led to a scarcity of training data, severely hindering the performance of automatic reaction extraction. In this work, we conduct a systematic study of active learning for chemical reaction extraction. We integrate six uncertainty- and diversity-based strategies with pretrained transformer-CRF architectures, and evaluate them on product extraction and role labeling task. While several methods approach full-data performance with fewer labeled instances, learning curves are often non-monotonic and task-dependent. Our analysis shows that strong pretraining, structured CRF decoding, and label sparsity limit the stability of conventional active learning strategies. These findings provide practical insights for the effective use of active learning in chemical information extraction.

Detecting jailbreak behaviour in large language models remains challenging, particularly when strongly aligned models produce harmful outputs only rarely. In this work, we present an empirical study of output based jailbreak detection under realistic conditions using the JailbreakBench Behaviors dataset and multiple generator models with varying alignment strengths. We evaluate both a lexical TF-IDF detector and a generation inconsistency based detector across different sampling budgets. Our results show that single output evaluation systematically underestimates jailbreak vulnerability, as increasing the number of sampled generations reveals additional harmful behaviour. The most significant improvements occur when moving from a single generation to moderate sampling, while larger sampling budgets yield diminishing returns. Cross generator experiments demonstrate that detection signals partially generalise across models, with stronger transfer observed within related model families. A category level analysis further reveals that lexical detectors capture a mixture of behavioural signals and topic specific cues, rather than purely harmful behaviour. Overall, our findings suggest that moderate multi sample auditing provides a more reliable and practical approach for estimating model vulnerability and improving jailbreak detection in large language models. Code will be released.

We introduce a multi-agent framework intended to emulate parts of a quantitative research team and support equity factor research on large financial panel datasets. QRAFTI integrates a research toolkit for panel data with MCP servers that expose data access, factor construction, and custom coding operations as callable tools. It can help replicate established factors, formulate and test new signals, and generate standardized research reports accompanied by narrative analysis and computational traces. On multi-step empirical tasks, using chained tool calls and reflection-based planning may offer better performance and explainability than dynamic code generation alone.

Conditional depth execution routes a subset of tokens through a lightweight cheap FFN while the remainder execute the standard full FFN at each controlled layer. The central difficulty is gate training: the gate decision must propagate through many layers before it influences the language modeling (LM) loss, so the resulting gradients are weak and noisy. Auxiliary losses are commonly stacked to stabilise training, yet the interactions among them -- particularly between a predictive auxiliary and explicit score supervision -- have not been systematically compared under controlled conditions. We evaluate two gate designs under a 157.5M-parameter decoder-only model with controller-only training, 50% full-path budget, and 3-seed runs on a fineweb-edu subset. The MLP gate (G1) maps the current hidden state to a utility score; the JEPA-guided gate (G3) adds an action-conditional predictor that forecasts, in a low-dimensional latent space, the outcome of executing full vs. cheap per token, aligned against a fixed target head. Under the standard recipe with oracle-style utility regression and pairwise rank supervision (util/rank), G3 improves early-to-mid optimisation over G1 in 3/3 seeds (lower avg LM, faster threshold hits, ~10.3x lower grad norms), with 20k-step endpoint LM within a 0.005 heuristic reference. A key finding (ablation A3): jointly removing util/rank improves best/avg LM and threshold-hit speed in 3/3 seeds for both gates, and the early-to-mid advantage of G3 over G1 disappears. We trace this to an off-policy oracle label that assumes all subsequent layers execute full, whereas gated execution routes only a fraction through full -- making util/rank net-negative under the current recipe. Removing util/rank also cuts the training FLOPs proxy from ~1.53x to ~1.07x full-only (2.87h to 1.75h on a V100-32GB, ~39%). Conclusions are scoped to the studied regime.

The choice of visualisation in empirical performance analysis is not a neutral presentation decision but an analytical one: different graphical forms reveal different features of the same dataset, and reliance on any single type systematically conceals what the others expose. This paper presents and argues for a suite of five complementary visualisation tools; tempographs, histograms with spline-smoothed probability density functions, ridgeline plots, stacked bar charts, and combination charts. These are applied to bar-level beats-per-minute data from recordings of Beethoven's five piano and cello sonatas (Op.~5 Nos.1 and2; Op.~69; Op.~102 Nos.1 and2) spanning 1930--2012. Each tool is described formally, its analytical properties characterised, its implementation detailed in working Python and MATLAB code, and its specific contribution demonstrated on a worked example using two recordings of Op.~5 No.~1 (Casals/Horszowski 1930--39 and Isserlis/Levin 2012) separated by eight decades. A five-panel composite figure applies all five tools to the same two recordings simultaneously, making the complementarity argument concrete: the tempograph reveals moment-to-moment structural parallels invisible in aggregate statistics; the spline-smoothed histogram exposes bimodality and secondary peaks suppressed by binning artefacts; the ridgeline plot positions both recordings within the full distributional space; the stacked bar chart shows divergent sectional pacing concealed by identical movement means; and the combination chart integrates mean tempo, variability, and historical reference marks in a single view. The spline-CDF smoothing method, applied to histogram data via cubic spline interpolation with zero-slope boundary conditions, is presented as a novel contribution to the performance analysis toolkit. Full implementation code is publicly available.

Recent advances in agentic frameworks have enabled AI agents to perform complex reasoning and decision-making. However, evidence comparing their reasoning performance, efficiency, and practical suitability remains limited. To address this gap, we empirically evaluate 22 widely used agentic frameworks across three reasoning benchmarks: BBH, GSM8K, and ARC. The frameworks were selected from 1,200 GitHub repositories collected between January 2023 and July 2025 and organized into a taxonomy based on architectural design. We evaluated them under a unified setting, measuring reasoning accuracy, execution time, computational cost, and cross-benchmark consistency. Our results show that 19 of the 22 frameworks completed all three benchmarks. Among these, 12 showed stable performance, with mean accuracy of 74.6-75.9%, execution time of 4-6 seconds per task, and cost of 0.14-0.18 cents per task. Poorer results were mainly caused by orchestration problems rather than reasoning limits. For example, Camel failed to complete BBH after 11 days because of uncontrolled context growth, while Upsonic consumed USD 1,434 in one day because repeated extraction failures triggered costly retries. AutoGen and Mastra also exhausted API quotas through iterative interactions that increased prompt length without improving results. We also found a sharp drop in mathematical reasoning. Mean accuracy on GSM8K was 44.35%, compared with 89.80% on BBH and 89.56% on ARC. Overall, this study provides the first large-scale empirical comparison of agentic frameworks for reasoning-intensive software engineering tasks and shows that framework selection should prioritize orchestration quality, especially memory control, failure handling, and cost management.

Artificial Intelligence is increasingly introduced into systems engineering activities, particularly within requirements engineering, where quality assessment and validation remain heavily dependent on expert judgment. While recent AI tools demonstrate promising capabilities in analyzing and generating requirements, their role within formal systems engineering processes-and their alignment with established INCOSE criteria-remains insufficiently understood. This paper investigates the extent to which AI-based tools can support systems engineers in evaluating requirement quality, without replacing professional expertise. The research adopts a structured systems engineering methodology to compare AI-assisted requirement evaluation with human expert assessment. A controlled study was conducted in which system requirements were evaluated against established INCOSE ``good requirement'' criteria by both experienced systems engineers and an AI-based assessment tool. The evaluation focused on consistency, completeness, clarity, and testability, examining not only accuracy but also the decision logic underlying each assessment. Results indicate that AI tools can provide consistent and rapid preliminary assessments, particularly for syntactic and structural quality attributes. However, expert judgment remains essential for contextual interpretation, ambiguity resolution, and trade-off reasoning. Rather than positioning AI as a replacement for systems engineers, the findings support its role as a decision-support mechanism within the RE lifecycle. From a systems engineering perspective, this study contributes empirical evidence on how AI can be integrated into RE workflows while preserving traceability, accountability, and engineering consistency.

In-context learning (ICL) allows large language models (LLMs) to adapt to new tasks from a few examples, making it promising for languages underrepresented in pre-training. Recent work on many-shot ICL suggests that modern LLMs can further benefit from larger ICL examples enabled by their long context windows. However, such gains depend on careful example selection, and the inference cost can be prohibitive for low-resource language communities. In this paper, we present an empirical study of many-shot ICL for machine translation from English into ten truly low-resource languages recently added to FLORES+. We analyze the effects of retrieving more informative examples, using out-of-domain data, and ordering examples by length. Our findings show that many-shot ICL becomes more effective as the number of examples increases. More importantly, we show that BM25-based retrieval substantially improves data efficiency: 50 retrieved examples roughly match 250 many-shot examples, while 250 retrieved examples perform similarly to 1,000 many-shot examples. We further show that ICL provides additional gains on top of fine-tuning.

Continuous knowledge updating for pre-trained large language models (LLMs) is increasingly necessary yet remains challenging. Although inference-time methods like In-Context Learning (ICL) and Retrieval-Augmented Generation (RAG) are popular, they face constraints in context budgets, costs, and retrieval fragmentation. Departing from these context-dependent paradigms, this work investigates a parametric approach using Low-Rank Adaptation (LoRA) as a modular knowledge memory. Although few recent works examine this concept, the fundamental mechanics governing its capacity and composability remain largely unexplored. We bridge this gap through the first systematic empirical study mapping the design space of LoRA-based memory, ranging from characterizing storage capacity and optimizing internalization to scaling multi-module systems and evaluating long-context reasoning. Rather than proposing a single architecture, we provide practical guidance on the operational boundaries of LoRA memory. Overall, our findings position LoRA as the complementary axis of memory alongside RAG and ICL, offering distinct advantages.

From generating headlines to fabricating news, the Large Language Models (LLMs) are typically assessed by their final outputs, under the safety assumption that a refusal response signifies safe reasoning throughout the entire process. Challenging this assumption, our study reveals that during fake news generation, even when a model rejects a harmful request, its Chain-of-Thought (CoT) reasoning may still internally contain and propagate unsafe narratives. To analyze this phenomenon, we introduce a unified safety-analysis framework that systematically deconstructs CoT generation across model layers and evaluates the role of individual attention heads through Jacobian-based spectral metrics. Within this framework, we introduce three interpretable measures: stability, geometry, and energy to quantify how specific attention heads respond or embed deceptive reasoning patterns. Extensive experiments on multiple reasoning-oriented LLMs show that the generation risk rises significantly when the thinking mode is activated, where the critical routing decisions are concentrated in only a few contiguous mid-depth layers. By precisely identifying the attention heads responsible for this divergence, our work challenges the assumption that refusal implies safety and provides a new understanding perspective for mitigating latent reasoning risks.

Robustness is widely viewed as a key challenge for real-world applications. However, because current research focuses only on difficult tasks, it partially captures real-world readiness. In this paper, we argue and verify that robustness, defined as consistency across semantically equivalent inputs, closely follows task difficulty: once models master a task, robustness emerges naturally. Through an empirical analysis of multiple models across diverse datasets and configurations (e.g., paraphrases, temperature changes), we observe a strong positive correlation between task performance and robustness. Furthermore, our findings indicate that robustness is driven primarily by task-specific competence rather than inherent model attributes, challenging the common view of robustness as an independent capability. This perspective implies that as tasks mature and model performance saturates, robustness on those tasks will similarly emerge. For researchers, this suggests that explicit efforts to measure robustness may deserve reduced emphasis, as robustness is likely to improve alongside performance. For practitioners, it signals that while many existing benchmarks are still unstable, models are already reliable on earlier tasks and suitable for deployment.

Social scientists are increasingly turning to unstructured datasets to unlock new empirical insights, e.g., estimating descriptive statistics of or causal effects on quantitative measures derived from text, audio, or video data. In many settings, unsupervised analysis is of primary interest, in that the researcher does not want to (or cannot) manually pre-specify all important aspects of the unstructured data to measure; they are interested in "discovery." This paper proposes a general and flexible framework for pursuing such discovery from unstructured data in a statistically principled way. The framework leverages recent methods from the literature on AI interpretability to map unstructured data points to high-dimensional, sparse, and interpretable "concept embeddings"; computes statistics from these concept embeddings for testing interpretable, concept-by-concept hypotheses; performs selective inference on these hypotheses using algorithms validated by new results in high-dimensional central limit theory, producing a selected set ("discoveries"); and both generates and evaluates human-interpretable natural language descriptions of these discoveries. The proposed framework has few researcher degrees of freedom, is robust to data snooping and other post-selection inference concerns, and facilitates fast and inexpensive sensitivity analysis and replication. Applications to recent descriptive and causal analyses of unstructured data in empirical economics are explored.

Object-Goal Navigation (ObjectNav) is a key capability for deploying mobile robots in everyday environments such as homes, schools, and workplaces. In this task, an agent must locate an instance of a target object category in previously unseen environments using only onboard perception, requiring the integration of semantic understanding, spatial reasoning, and long-horizon planning. Reinforcement learning (RL) has become a dominant paradigm for ObjectNav, yet modern systems involve numerous design choices across perception modules, policy architectures, and inference-time strategies. The relative impact of these components, however, remains poorly understood. In this work, we present a large-scale empirical study of modular RL-based ObjectNav systems. We decompose the navigation pipeline into three key components: perception, policy, and test-time enhancement, and conduct extensive controlled experiments to analyze their individual contributions. Our results suggest that improvements in perception quality and test-time strategies often yield larger performance gains than policy improvements alone, highlighting the importance of understanding how different components interact within modular navigation systems. Motivated by these findings, we introduce a unified framework for systematically studying modular ObjectNav systems. Guided by our analysis, we build an enhanced system that achieves state-of-the-art performance on the Gibson benchmark, improving SPL by 6.6% and success rate by 2.7% over prior methods. We also introduce a human expert baseline, achieving 98% success, highlighting the significant gap between current RL agents and human-level navigation. Finally, we provide practical insights and design recommendations for each module to help guide future research. Project page: https://honwang0054.github.io/What-matters-in-RL-ObjNav-web/.

We propose persistent discrete homology as a tool for topological data analysis and discuss its advantages over the existing methods. In particular, we provide empirical evidence that persistent discrete homology is more noise-resistant than persistent homology of the Vietoris-Rips complex for data coming from non-metric settings.