Academic Intelligence · Curated Daily

Explore the Frontier of Global Academia

AcademicHub aggregates real-time literature from top journals and preprint platforms. Build your personal research radar and let large language models compile cross-disciplinary analysis briefings automatically.

01.
arXiv (math.PR) 2026-06-24

Variational Tail Bounds for Norms of Random Vectors and Matrices

arXiv:2503.17300v5 Announce Type: replace Abstract: We propose a variational tail bound for norms of random vectors and matrices under moment assumptions on their one-dimensional marginals. A simplified version of the bound that parametrizes the ``aggregating distribution'' using a certain pushforward of the Gaussian distribution is also provided. We apply the proposed method to reproduce some of the well-known bounds on norms of Gaussian random vectors, and also obtain dimension-free tail bounds for the Euclidean norm of random vectors with arbitrary moment profiles. Furthermore, we reproduce a dimension-free concentration inequality for sum of independent and identically distributed positive semidefinite matrices with sub-exponential marginals, and obtain a concentration inequality for the sample covariance matrix of sub-exponential random vectors. We also obtain a tail bound for the operator norm of a random matrix series whose random coefficients may have arbitrary moment profiles. Furthermore, we use coupling to formulate an abstraction of the proposed approach that applies more broadly. As a corollary, we derive a PAC-Bayesian-style bound in terms of a certain combination of the KL and R\'{e}nyi divergences between the prior and posterior distributions.

02.
arXiv (CS.CL) 2026-06-24

Selective Rotary Position Embedding

Position information is essential for language modeling. In softmax transformers, Rotary Position Embeddings (RoPE) encode positions through fixed-angle rotations, while in linear transformers, order is handled via input-dependent (selective) gating that decays past key-value associations. Selectivity has generally been shown to improve language-related tasks. Inspired by this, we introduce Selective RoPE, an input-dependent rotary embedding mechanism, that generalizes RoPE, and enables rotation in arbitrary angles for both linear and softmax transformers. We show that softmax attention already performs a hidden form of these rotations on query-key pairs, uncovering an implicit positional structure. We further show that in state-space models and gated linear transformers, the real part manages forgetting while the imaginary part encodes positions through rotations. We validate our method by equipping gated transformers with Selective RoPE, demonstrating that its input-dependent rotations improve performance in language modeling and on difficult sequence tasks like copying, state tracking, and retrieval.

03.
arXiv (CS.CV) 2026-06-17

Gaussian Light Field Splatting: A Physical Prior-Driven Vision Transformer for Unsupervised Low-Light Image Enhancement

Existing unsupervised low-light image enhancement methods often encounter local exposure imbalance and color distortion under complex non-uniform illumination. In addition, most Vision Transformers lack an explicit mechanism for modeling the physical priors of illumination degradation. To address these limitations, we propose GLFS, a Gaussian light field splatting-based Vision Transformer that integrates continuous physical illumination modeling from Gaussian splatting into the Transformer architecture. In GLFS, scene illumination is represented by a superposition of anisotropic Gaussian basis functions. Physics-guided biases are introduced into self-attention to adaptively infer a spatial gain field, enabling accurate and uniform restoration under complex illumination. To reduce color bias and structural degradation during enhancement, a color-vector angular loss and a luminance-edge loss are further developed. These losses enforce hue consistency and improve the structural fidelity of local details. Extensive ablation studies and quantitative evaluations show that GLFS provides clear advantages in illumination correction and detail preservation. It achieves state-of-the-art performance and offers a new representation paradigm for low-light image enhancement.

04.
medRxiv (Medicine) 2026-06-22

Demographic Calibration Gaps in Breast Cancer Risk Prediction: Introducing the Demographic Calibration Gap Score

Authors:

ABSTRACT: Most breast cancer prediction studies skip calibration reporting entirely. Fewer still examine calibration by demographic subgroup. Predicted probabilities that are systematically off for specific racial or gender groups produce biased clinical decisions, and aggregate statistics will not catch that. Objective: To introduce the Demographic Calibration Gap Score (DCGS), a metric that measures how much calibration error varies across demographic subgroups, and to show how it performs across five classifiers, four calibration conditions, and two datasets. Methods: Five classifiers were trained on the Wisconsin Diagnostic Breast Cancer dataset (n=569) and evaluated on a breast cancer cohort from MIMIC-IV (n=1,316). Three global calibration methods were applied: no calibration, Platt scaling, and isotonic regression. A fourth condition, subgroup-targeted Platt scaling, was applied to the MIMIC cohort. DCGS was computed as across racial and gender subgroups, with 95% bootstrap confidence intervals. Conformal prediction coverage and Demographic Coverage Gap (DCG) were reported. Results: On Wisconsin, all five models achieved AUROC above 0.98 and ECE below 0.12. Performance fell sharply on the MIMIC external cohort: AUROC dropped to 0.45-0.57 for base and globally calibrated variants, confirming distributional shift. DCGS exceeded the 0.05 clinical significance threshold in 28 of 40 model-calibration combinations on the race axis. Neither global Platt nor isotonic calibration reliably reduced DCGS below that threshold. Conformal coverage collapsed to roughly 25% on MIMIC, and racial DCG exceeded 0.15 for all 20 model-variant combinations. Conclusions: Reducing population-level ECE through global recalibration does not reliably close demographic calibration gaps. DCGS gives researchers a direct, standardized way to detect and report those disparities. Code and the DCGS computation library are released as open-source Python under the MIT License.

05.
arXiv (CS.CV) 2026-06-11

MB-Loc: Multi-planar Bird's-eye-view Localization in outdoor LiDAR scenes

Global LiDAR localization is a fundamental task for autonomous navigation systems. Recent methods perform Scene Coordinate Regression (SCR) and achieve superior accuracy over Absolute Pose Regression (APR) solutions by predicting dense 3D world coordinates. However, SCR approaches introduce two major bottlenecks: severe computational inefficiency from processing raw 3D geometries and significant performance degradation under varying sensor viewpoints. To address these limitations, we present MB-Loc, a lightweight and viewpoint-robust SCR framework. Instead of relying on heavy 3D convolutions, we project the input LiDAR scan into a 2.5D Multi-planar Bird's-Eye View (BEV) representation. By slicing the point-cloud along the Z-axis and mapping signed depths into discrete 2D planes, MB-Loc retains essential 3D geometric structures while exploiting the computational tractability of standard 2D CNNs. To handle the inherent sparsity of outdoor LiDAR, we introduce a KL-regularized latent bottleneck that explicitly models spatial uncertainty without injecting stochastic noise. Finally, to ensure rotation robustness, we apply 3D spatial augmentations prior to planar projection, forcing the network to implicitly learn viewpoint-invariant features. We perform extensive experiments on the publicly available NCLT dataset and demonstrate that our proposed method outperforms the current state-of-the-art. Operating at real-time inference speeds, MB-Loc significantly outperforms traditional 3D-SCR architectures in computational efficiency.

06.
arXiv (math.PR) 2026-06-16

Flowing to Normality and the Fate of the Single Ring Theorem

arXiv:2606.15791v1 Announce Type: cross Abstract: Random non-hermitian matrix ensembles with double-sided rotation invariance obey, in the limit of large matrix size, the Single Ring Theorem, which states that the support of the mean eigenvalue distribution in the complex plane is either a disk or an annulus. In contrast, rotational-invariant random normal matrix ensembles can have mean eigenvalue densities supported over any number of concentric annuli in the complex plane. In this paper we introduce and investigate, both analytically and numerically, a non-hermitian matrix model which flows from a generic matrix distribution obeying the Single Ring Theorem to a distribution of normal matrices by tuning a parameter which penalizes non-normality. We observe numerically breakdown of the Single Ring Theorem as the model flows towards normality, and determine the critical value of the parameter at which the transition occurs. We also study in detail the behavior of the singular values of these matrices under the flow. These singular values form a Fermi gas confined to the positive half-line. In particular, we find that at small values of the flow parameter, the interparticle spacings in the gas exhibit Wigner-Dyson repulsion, whereas for asymptotically large values of the flow parameter, at the normal matrix endpoint of the flow, the spacing statistics is Poissonian. The flow interpolates continuously between these two types of statistics. However, this change in statistics is not related directly to breaking of the Single Ring Theorem, which occurs very early-on along the flow, in the regime of Wigner-Dyson statistics. Finally, we introduce a certain ensemble of random permutations associated with the gas, and make a conjecture on how to use it in order to reconstruct approximately the average density of complex eigenvalues from that of the singular values in the large-$N$ limit.

07.
arXiv (CS.CL) 2026-06-24

AGORA: An Archive-Grounded Benchmark for Agentic Workplace Document Reasoning

Large language models are increasingly deployed as agents that reason over documents rather than answer from parametric knowledge. We study archive-grounded reasoning: locating sparse evidence across a large, messy collection of workplace files, reconciling inconsistent terminology, units, and time conventions, and computing an answer. Existing benchmarks address only parts of this setting and none jointly stresses archive-groundedness, agentic exploration, and cross-domain coverage. We introduce Agora, a benchmark pairing 362 questions with eight domain collections of 9,664 authentic documents and 372M tokens, far exceeding any model's context window, so agents must explore deliberately rather than scan exhaustively. Agora is built by an agentic pipeline combining cross-document task synthesis, leakage-preventing obfuscation, and difficulty filtering. Evaluating eight models, we find the task far from solved: even the strongest reaches only 59.4% accuracy, with notable variation across domains.

08.
arXiv (CS.LG) 2026-06-19

Probe-and-Refine Tuning of Repository Guidance for Coding Agents

arXiv:2606.20512v1 Announce Type: cross Abstract: LLM-based coding agents need higher-level operational knowledge about a repository (which files house which subsystems, how to run the test suite, which workflows have historically led to wrong fixes) that does not exist in the code itself. Engineers typically maintain \texttt{AGENTS.md} files to supply this context as instructions for coding agents, but whether they help is contested: recent studies disagree on whether LLM-generated guidance improves or harms agent performance. In this paper we show that how the guidance is produced is the decisive variable, and introduce probe-and-refine tuning: a procedure that uses synthetic bug-fix probes to iteratively diagnose and patch a repository's guidance file through single-shot LLM calls, with no agent loop or tool use during tuning. On SWE-bench Verified across four independent trials with Qwen3.5-35B-A3B at 200 steps, probe-and-refine achieves 33.0\,\% mean resolve rate vs.\ 28.3\,\% for the static knowledge base used to initialize it and 25.5\,\% for an unguided baseline ($p < 0.001$ for both probe-and-refine contrasts). The improvement comes from coverage rather than precision: refined guidance produces evaluable patches for 14.5 percentage points (pp) more instances while per-patch precision remains statistically constant ($\sim$59\,\%, $p = 0.119$), showing that improved guidance helps agents reach the correct file rather than improving the quality of the changes they make. Further, a step-budget experiment shows that guidance is what lets the agent use a larger step budget productively, and a cross-model experiment with NVIDIA-Nemotron-3-Nano-30B-A3B finds that the tuning loop degrades when the model cannot generate sufficiently diagnostic output, though per-patch precision remains constant even then.

09.
arXiv (CS.AI) 2026-06-16

The Integrator Advantage: Controlled Agentic AI for Small and Medium-Sized Companies

arXiv:2606.16649v1 Announce Type: new Abstract: Agentic AI marks a new phase of enterprise automation. Unlike traditional automation or conversational AI, agentic systems can interpret goals, plan multi step tasks, access tools, interact with enterprise systems, and execute workflows with varying degrees of autonomy. For small and medium sized companies, this creates potential to reduce administrative burden, accelerate routine processes, and improve the use of organizational knowledge. This paper argues that the near term value of Agentic AI does not lie in full autonomy or workforce reduction, but in controlled partial autonomy for simple and medium complexity business processes. It proposes an integration framework covering use case suitability, autonomy levels, technical integration, governance, security, employee enablement, and measurable impact. The paper concludes that Agentic AI can become a productivity lever when implemented as a human centered capability with responsibility and accountability retained by people.

10.
arXiv (CS.LG) 2026-06-17

Memory-Efficient Meta-Reinforcement Learning for Adaptive Safety-Critical Control in Adversarial Spacecraft Proximity Operations

arXiv:2606.17414v1 Announce Type: new Abstract: Autonomous spacecraft rendezvous and proximity operations (RPO) require controllers that guarantee safety under thrust constraints while minimizing fuel expenditure. Input-constrained control barrier functions (ICCBFs) provide a control method for nonlinear systems with actuation constraints that construct a forward-invariant safe set. Previous work has shown that learning class-$\mathcal{K}$ functions defining the ICCBF recursion via meta reinforcement learning (meta-RL) yields a robust, non-greedy approach to safety-critical control in RPO. This paper extends that framework further by investigating the performance of three recurrent network architectures (Long Short Term Memory (LSTM), Gated Recurrent Unit (GRU), Selective State Space Model (Mamba)) and two training algorithms (Proximal Policy Optimization (PPO) and Soft Actor Critic (SAC)) to identify the best setup for tuning ICCBF class-K functions via meta-RL. In addition to cooperative test cases, performance is evaluated in the presence of adversarial behavior where the target spacecraft behaves in a way that worsens the safety of the chaser spacecraft. Results indicate that state space models such as Mamba when used with PPO achieve superior task completion, safety, and fuel-savings compared to other architectures, across all cooperative and uncooperative scenarios tested.

11.
arXiv (CS.LG) 2026-06-16

Enhancing Physics-Informed Neural Networks Through Feature Engineering

arXiv:2502.07209v4 Announce Type: replace Abstract: Physics-Informed Neural Networks (PINNs) seek to solve partial differential equations (PDEs) with deep learning. Mainstream approaches that deploy fully-connected multi-layer deep learning architectures require prolonged training to achieve even moderate accuracy, while recent work on feature engineering allows higher accuracy and faster convergence. This paper introduces SAFE-NET, a Single-layered Adaptive Feature Engineering NETwork that achieves orders-of-magnitude lower errors with far fewer parameters than baseline feature engineering methods. SAFE-NET returns to basic ideas in machine learning, using Fourier features, a simplified single hidden layer network architecture, and an effective optimizer that improves the conditioning of the PINN optimization problem. Numerical results show that SAFE-NET converges faster and typically outperforms deeper networks and more complex architectures. It consistently uses fewer parameters – on average, 65% fewer than the competing feature engineering methods – while achieving comparable accuracy in less than 30% of the training epochs. Moreover, each SAFE-NET epoch is 95% faster than those of competing feature engineering approaches. These findings challenge the prevailing belief that modern PINNs effectively learn features in these scientific applications and highlight the efficiency gains possible through feature engineering.

12.
arXiv (CS.AI) 2026-06-11

DataEvolver: Automatic Data Preparation for Large Language Models through Multi-Level Self-Evolving

arXiv:2606.07001v2 Announce Type: replace-cross Abstract: High-quality training data is essential to large language models (LLMs) and typically requires extensive and costly manual curation. Existing automatic data preparation methods rely on predefined pipelines or customized human instructions, which limits their adaptability to diverse data distributions and lacks principled guidance from high-quality examples. In this paper, we introduce DataEvolver, the first self-evolving data preparation system that automatically constructs pipelines to transform raw data into high-quality data. DataEvolver employs a multi-level mechanism to ensure both pipeline executability and effectiveness. At the operator level, it incrementally expands the operator set to construct a logical plan while resolving dependency conflicts. At the pipeline level, it instantiates logical plans into executable code and iteratively refines pipeline orchestration through a feedback loop that reduces the distribution gap between prepared data and high-quality examples. Experiments on seven benchmarks show that DataEvolver substantially improves data quality and achieves an average 10\% gain in downstream LLM performance compared with training on original data, highlighting new opportunities for the iterative co-evolution of LLMs and data.

13.
arXiv (CS.LG) 2026-06-16

Peak-Based Nuclide Identification in HPGe $\gamma$-Spectrometry with Machine Learning and SHAP

arXiv:2606.14874v1 Announce Type: cross Abstract: High-purity germanium gamma spectra often require time-consuming analyses from subject matter experts. Photopeaks within these spectra are carefully fitted and numerical methods are employed to assist with nuclide identification (NID) and quantification. Amending the list of nuclides identified by analysis software can be nontrivial. When many samples need to be analyzed, it is therefore challenging to make timely and correct decisions. Supervised machine-learning-based NID can serve as an expert-informed, automated tool to improve the initial set of radionuclides suggested to an analyst and more effectively drive subsequent quantification. To that end, we implemented machine learning models that map photopeaks carefully fitted by analysts to NID results for experimental spectra containing various isotopic combinations drawn from a set of 65 isotopes. The best model achieved an F1 score of 0.97, markedly surpassing the F1 score of 0.84 achieved by traditional software when compared using a nuclide library comprising the same 65 isotopes assessed by the models. Finally, we illustrated the most important input features for model predictions using Shapley Additive Explanations. These explanations revealed that the models use physically relevant photopeaks when making predictions for the isotopes in our nuclide library.

15.
medRxiv (Medicine) 2026-06-24

Cognitive and Neuroimaging Biomarker Intra-Individual Variability in Alzheimer's Disease

Background Greater cognitive intra-individual variability (IIV) reflects increased heterogeneous performance across cognitive domains and has been linked to a higher risk of Alzheimer's disease (AD). However, it remains unclear whether cognitive IIV is linked to heterogeneous dispersion of regional AD pathology. Hence, we aimed to examine the association between cognitive IIV and AD neuroimaging biomarker IIV. Methods This study included participants with normal cognition (CN) and mild cognitive impairment (MCI) from the Alzheimer's Disease Neuroimaging Initiative. Cognitive IIV was computed as the within-person standard deviation of five domain-specific neuropsychological test z-scores. Four neuroimaging biomarker IIV metrics were similarly derived using regional amyloid-{beta} (n = 1,021), tau (n = 719), cortical thickness (n = 2,148), and combined amyloid-tau-neurodegeneration (ATN, n = 258). Associations between cognitive IIV and each biomarker IIV were evaluated using linear regression models, adjusted for relevant covariates. Results Higher cognitive IIV was associated with greater biomarker IIV across amyloid-{beta} ({beta} = 0.039, SE = 0.014, p = .006), tau ({beta} = 0.196, SE = 0.033, p < .001), cortical thinning ({beta} = 0.036, SE = 0.008, p < .001), and ATN ({beta} = 0.176, SE = 0.043, p < .001). Interaction analyses revealed that the associations of cognitive IIV with tau IIV, cortical thickness IIV, and ATN IIV were stronger in MCI than CN individuals. Significant interactions between cognitive IIV and biomarker positivity status showed that the effect with amyloid-{beta} IIV was attenuated in A- ({beta} = 0.004, SE = 0.014, p = .78) but that the effect with tau IIV remained robust even in T- individuals ({beta} = 0.088, SE = 0.022, p < .001). Conclusion Elevated cognitive IIV is associated with greater heterogeneity in cortical dispersion of AD-related pathology, particularly in prodromal AD and in the presence of abnormal pathology. As a novel measure that captures variation in topographical scattering of AD pathological burden across the cortex, AD biomarker IIV may offer research and clinical utility beyond evaluating absolute biomarker load or thresholds.

16.
arXiv (CS.CV) 2026-06-19

CUPID: Reconstructing UV Texture Maps for Interpretable Person-of-Interest Deepfake Detection

Deepfakes targeting a high-profile individual, known as Person-of-Interest (POI), are a threat to modern democracies and societies. Current POI deepfake detection methods still struggle to combine robustness to post-processing, efficiency and interpretability, focal aspects of modern deepfake detectors. In this paper we propose CUPID, a POI video deepfake detector that combines UV texture maps, a facial appearance representation derived from 3D face reconstructions, with the representation learning capabilities of the Masked Autoencoder (MAE). Our method does not require any deepfake videos in its training phase. Moreover, it does not even require to include a specific POI in the training set: the combination of UV texture maps extracted from real video frames and the MAE context-guided reconstruction yields a latent space that captures rich and discriminative facial features also for identities unseen during training. In the testing phase, the embeddings extracted from a query video depicting the POI can be matched against pristine reference videos to assess the video authenticity. Furthermore, operating in the UV space naturally provides an additional layer of interpretability. Specifically, we can extract decoded residual maps that highlight which facial regions of a test video deviate most from the identity representation of the corresponding POI. Experiments on four deepfake datasets show that CUPID outperforms current state of the art on most datasets and achieves the best overall robustness against strong downscaling and compression, providing also substantially faster inference. Our experimental code will be released at https://github.com/polimi-ispl/CUPID.

17.
PLOS Medicine 2026-06-01

Prenatal exposure to asthma medications and risk of neurodevelopmental disorders and educational difficulties: A systematic review and meta-analysis

by Lama A. Shakhshir, Alexia Karain, Jill P. Pell, Claire E. Hastie, Scott M. Nelson, Michael Fleming Background Since asthma exacerbations during pregnancy risk maternal and fetal health, continued medication is important. However, some studies have reported adverse neurodevelopmental outcomes following prenatal exposure to asthma medication. Therefore, this systematic review aimed to collate the existing evidence on the associations between prenatal exposure to asthma medication and neurodevelopmental and educational outcomes. Methods and findings A systematic review was conducted in accordance with PRISMA guidelines and the PECO framework. PubMed, Medline and Embase databases were searched for studies investigating prenatal exposure to one or more asthma medication and neurodevelopmental or educational outcomes published, in English, between January 2003 and September 2024, and updated in November 2025. Studies of asthma medication used for other indications were excluded. Study quality was assessed using the Newcastle-Ottawa scale. Random-effects meta-analyses were conducted where appropriate and heterogeneity was evaluated using Cochran’s Q and I2 tests.Of 16,824 studies identified by the initial search, seven were eligible for inclusion. All investigated beta-2-adrenergic agonists (B2AA), with one including B2AA as mono- and polytherapy—and one study also investigated inhaled corticosteroids (ICS) exposure. Two reported associations with autism spectrum disorder (ASD) and one with attention-deficit hyperactivity disorder (ADHD). An updated search identified one additional eligible study, which examined both ADHD and ASD, as well as other neurodevelopmental disorders. The included eight studies (n = 3,867,170 participants) comprised cohort (n = 5) and case-control (n = 3) designs and reported inconsistent results. Meta-analysis of three studies (n = 1,380,871) indicated significant associations with ASD for exposure to B2AA both preconception (aOR 1.34, 95% CI [1.19,1.52]) and during pregnancy (aOR 1.29, 95% CI [1.16,1.42]). Heterogeneity was low, with no evidence of significant publication bias. Limitations of the included studies comprised residual confounding and exposure misclassification. Additionally, studies included in the meta-analysis were few in number and did not adequately distinguish between medication effects and underlying maternal asthma. Conclusion Meta-analysis suggested an association between prenatal exposure to B2AA and ASD. An association with ADHD, reported in a single study, requires corroboration. To date, based on our search strategy, no association has been reported with communication skills, motor skills, problem-solving and personal-social skills, or cerebral palsy.

18.
medRxiv (Medicine) 2026-06-15

Modelling the public-health impact of indoor air quality interventions on respiratory virus transmission

Respiratory virus transmission occurs in indoor settings where ventilation, occupancy, and dwell time determine exposure levels. Improving indoor air quality (IAQ) therefore could help reduce disease burden associated with respiratory viruses, yet its population-level impact remains poorly quantified. Here, we develop an individual-based transmission modelling framework that links within-location airborne dynamics to individual infection risk and population-level spread, whilst explicitly incorporating heterogeneity in ventilation and baseline indoor air quality across locations. We use this modelling approach to evaluate IAQ-improving interventions (air-quality interventions or AQIs), using hypothetical endemic and pandemic pathogen archetypes with properties similar to SARS-CoV-2 and influenza, and evaluate how effects on key epidemiological metrics (such as annualized incidence and epidemic final size) depend on AQI coverage, efficacy and allocation strategy. At 20% AQI intervention coverage and 80% efficacy, annualized incidence was reduced by approximately 7.2% for an endemic 'SARS-CoV-2-like' respiratory virus, and 17.0% for an endemic 'influenza-like' virus; at 60% coverage (80% efficacy) the reductions were 26.3% and 56.4%, respectively. Targeting AQI installation to the highest-risk locations outperformed random allocation: for SARS-CoV-2-like transmission, 20% coverage at 80% efficacy cut absolute incidence by 10.8% when targeted versus 7.2% when random; for influenza-like transmission, this comparison was 28.9% versus 17.0%. In epidemic scenarios, random installation at 40% coverage and 60% efficacy reduced final size by 23.7% (influenza-like) versus 6.3% (SARS-CoV-2-like). These results support treating clean indoor air as core public-health infrastructure and prioritising risk-based deployment of IAQ-improving interventions to maximise population-level benefit within budgetary and operational constraints.

19.
arXiv (CS.LG) 2026-06-11

Bergson: An Open Source Library for Data Attribution

arXiv:2606.11660v1 Announce Type: new Abstract: Data attribution is a promising field in interpretability that aims to explain model behavior through the influence of its training data, with applications including debugging undesirable model behavior and training dataset curation. However, significant engineering effort is required to perform it at scale, and many cutting edge techniques lack open-source tooling and support. Bergson is an open source library that aims to enable faster progress in the field by providing a host of techniques that scale to very large language models and pre-training datasets. The library natively supports on-disk gradient stores and multi-node distributed training, and provides quality of life tools for researchers. Finally, we introduce the first open-source implementations of three leading data attribution methods: MAGIC, SOURCE, and TrackStar. The library is available at https://github.com/EleutherAI/bergson .

20.
arXiv (CS.AI) 2026-06-12

Examining the Usage of Generative AI Models in Student Learning Activities for Software Programming

arXiv:2511.13271v2 Announce Type: replace-cross Abstract: The rise of Generative AI (GenAI) tools like ChatGPT has created new opportunities and challenges for computing education. Existing research has primarily focused on GenAI's ability to complete educational tasks and its impact on student performance, often overlooking its effects on knowledge gains. In this study, we investigate how GenAI assistance compares to conventional online resources in supporting knowledge gains across different proficiency levels. We conducted a controlled user experiment with 24 undergraduate students of two different levels of programming experience (beginner, intermediate) to examine how students interact with ChatGPT while solving programming tasks. We analyzed task performance, conceptual understanding, and interaction behaviors. Our findings reveal that generating complete solutions with GenAI significantly improves task performance, especially for beginners, but does not consistently result in knowledge gains. Importantly, usage strategies differ by experience: beginners tend to rely heavily on GenAI toward task completion often without knowledge gain in the process, while intermediates adopt more selective approaches. We find that both over-reliance and minimal use result in weaker knowledge gains overall. Based on our results, we call on students and educators to adopt GenAI as a learning rather than a problem solving tool. Our study highlights the urgent need for guidance when integrating GenAI into programming education to foster deeper understanding.

21.
arXiv (CS.CL) 2026-06-16

Do You Really Need a GPU to Guard Your LLM? CPU-Class Classifiers and Multi-Stage Pipelines for Safety Enforcement at Scale

Safety classifiers that screen LLM inputs for jailbreak attempts have become standard deployment components, yet almost all production systems rely on GPU-based models: fine-tuned transformers and LLM-as-a-judge pipelines. These approaches impose significant per-query latency and infrastructure cost. Very little research has asked whether CPU-based classifiers, such as support vector machines and gradient-boosted trees trained on TF-IDF features, can match their accuracy across the conditions that production deployments encounter. We evaluate five CPU classifier families, Mamba-130M as an SSM-based GPU classifier, and transformer-based GPU models (DeBERTa-v3 and Gemma-2B with LoRA) across nine jailbreak sources and three regimes: in-distribution (D1), out-of-distribution (D2), and adversarially obfuscated (D3). On D1, the best CPU classifier matches the best transformer GPU model at roughly one-fifth the deployment cost. On D2, CPU classifiers fail via confident miscalibration, producing high-confidence false negatives that bypass escalation entirely. On D3, CPU classifiers outperform transformer GPU models by more than 26 percentage points in F1. Based on these complementary failure modes, we design GuardChain, a three-stage safety pipeline (Regex -> CPU -> GPU) that routes each prompt to the cheapest stage capable of a confident decision. The CPU stage alone resolves 80\% of in-distribution prompts at near-peak accuracy, and the GPU stage recovers the out-of-distribution failures. For practitioners deploying LLM safety at scale, this work provides evidence that GPU-class infrastructure is unnecessary for the majority of traffic.

22.
arXiv (CS.AI) 2026-06-17

Quantifying Consistency in LLM Logical Reasoning via Structural Uncertainty

arXiv:2606.17312v1 Announce Type: new Abstract: Large language models can arrive at the same answer through reasoning paths that are unstable, contradictory, or difficult to rank consistently – a failure mode especially prevalent in multi-step deductive reasoning. Existing methods assess reliability primarily through output dispersion – measuring how much sampled answers differ – but this discards a complementary signal: whether the model can consistently rank competing reasoning candidates. We propose structural uncertainty, a consistency-aware framework derived from the stability of self-preference-induced rankings over sampled reasoning solutions. Given a query, we generate multiple candidate solutions and ask the model to judge pairwise preferences among its own outputs. We aggregate self-preferences into ranking distributions via Bradley-Terry modeling with PageRank, and decompose the signal into two entropy-based components: across-trial ranking instability and within-trial candidate ambiguity. Across five LLMs and eight benchmarks, structural signals provide information complementary to answer dispersion: on logical and mathematical reasoning tasks, the combination improves identification of unreliable instances, while on factual retrieval the structural signal collapses toward uniformity, diagnosing a regime boundary where reasoning-level consistency evaluation is uninformative. The two components relate differently to accuracy: within-trial ambiguity correlates positively with correctness – consistent with settings where multiple plausible solution paths remain competitive – while across-trial instability correlates negatively, signaling unreliable reasoning. Structural uncertainty is best understood not as a universal confidence estimator, but as a regime-sensitive evaluator of logical reasoning consistency.

23.
arXiv (CS.AI) 2026-06-16

Discrete optimal transport is a strong audio adversarial attack

arXiv:2509.14959v3 Announce Type: replace-cross Abstract: In this paper, we investigate discrete optimal transport (DOT) as a black-box attack against modern automatic speaker verification (ASV) and anti-spoofing countermeasure (CM) systems. Our attack operates as a post-processing distribution-alignment step. Frame-level WavLM embeddings of generated speech (or another person speech) are aligned to an unpaired bona fide speech pool using entropic optimal transport and a top-k barycentric projection, followed by neural vocoding. Unlike gradient-based attacks, the proposed method requires no access to model parameters, gradients, or training data. Experiments on ASVspoof2019 and ASVspoof5 demonstrate that DOT attack substantially increases CM EER and substantially degrades ASV performance across multiple spoofing attacks. The attack transfers across datasets and remains effective after CM fine-tuning. Analysis using speaker similarity, Fréchet Audio Distance, and visualization of embedding distributions suggests that DOT succeeds by shifting source speech toward bona fide regions of the representation space rather than by maximizing speaker similarity. These results indicate that optimal-transport-based distribution alignment represents a previously underexplored attack vector for contemporary ASV and anti-spoofing systems.

24.
arXiv (CS.LG) 2026-06-11

Counterexample Guided Learning in the Large using Reasoning Agents

arXiv:2606.11521v1 Announce Type: new Abstract: LLMs and LLM agents should improve when given feedback, but identifying when they are able to do so is difficult: feedback is heterogeneous, domain-specific, and difficult to control. We approach this challenge by asking LLMs to perform regular-expression induction, a classical symbolic learning problem where precise mechanisms for feedback exist in the form of counterexamples. In counterexample-guided learning, a learner (LLM) proposes candidate regular expressions from positive/negative-labeled strings, and the teacher (verifier) returns counterexamples showcasing the difference between the candidate and target languages. We identify novel counterexample-guided refinement strategies that enable effective regex learning, such as regularization and symbolic counterexample clusters. We also explore agentic strategies such as reflection and repair loops. Empirically, we find that verifier feedback substantially improves sample efficiency on challenging regex-induction tasks, reducing the number of labeled examples required and enabling learning of complex target expressions where standard prompting fails. For example, on the hardest task groups, our counterexample-guided framework improves success from 3.2% to 38.1% and from 38.9% to 74.1% on two different regex domains. These results suggest that LLMs can benefit from rich feedback beyond treating it as additional data, opening the door for robust verifier-guided methods for LLM-based program synthesis and formal reasoning.

25.
arXiv (CS.AI) 2026-06-18

SciRisk-Bench: A Risk-Dimension-Aware Benchmark for AI4Science Safety

arXiv:2606.18936v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly embedded in AI for Science (AI4Science) workflows, from scientific question answering and literature analysis to laboratory planning and autonomous discovery. This progress creates an urgent need for safety benchmarks that evaluate not only scientific competence, but also whether models recognize and avoid risks in high-stakes scientific contexts. Existing AI4Science safety datasets cover several disciplines and task formats, leaving the underlying risk dimensions underspecified. We introduce SciRisk-Bench, a benchmark designed to evaluate AI4Science safety from two complementary perspectives: explicit risk dimensions and scientific disciplines. SciRisk-Bench covers 7 disciplines, 31 subdisciplines and 10 risk dimensions. In the experimental section, we evaluate both mainstream LLMs and science-oriented LLMs across risk dimensions, disciplines, and sub-disciplines, enabling fine-grained diagnosis of where scientific models remain unsafe.