# Generative Adversarial Network

A generative adversarial network is a neural-network model involving a generator and discriminator trained in an adversarial minimum-maximum game.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## Density Feedback Control for Transferability of Generative Sparse Adversarial Attacks

DevFeed: [Density Feedback Control for Transferability of Generative Sparse Adversarial Attacks](<https://devfeed.tech/articles/density-feedback-control-for-transferability-of-generative-sparse-adversarial-attacks-145000.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.25001>)

Author: Dongsu Song, Jay Hoon Jung

Published: 2026-11-20T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Adversarial attacks](<https://devfeed.tech/topics/adversarial-attacks.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [adversarial](<https://devfeed.tech/tags/adversarial.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [asr](<https://devfeed.tech/tags/asr.md>), [attack](<https://devfeed.tech/tags/attack.md>), [attacks](<https://devfeed.tech/tags/attacks.md>), [black-box](<https://devfeed.tech/tags/black-box.md>), [cnn](<https://devfeed.tech/tags/cnn.md>), [control](<https://devfeed.tech/tags/control.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [decoder](<https://devfeed.tech/tags/decoder.md>), [encoder](<https://devfeed.tech/tags/encoder.md>), [encoder-decoder](<https://devfeed.tech/tags/encoder-decoder.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [feature](<https://devfeed.tech/tags/feature.md>), [feature-importance](<https://devfeed.tech/tags/feature-importance.md>), [feedback](<https://devfeed.tech/tags/feedback.md>), [fixed](<https://devfeed.tech/tags/fixed.md>), [gap](<https://devfeed.tech/tags/gap.md>), [generative](<https://devfeed.tech/tags/generative.md>), [generator](<https://devfeed.tech/tags/generator.md>)

### AI overview

Density Feedback Control (DFC) adjusts the sparsity coefficient after each training epoch based on the difference between observed hard-mask density and a target, without changing the generator architecture or loss. At a 10% target, it brought three base attacks' mean density within one percentage point in a single run. Two attacks improved transfer attack success rates on CNN and Transformer targets, including under per-image ℓ₀ budgets, while perturbing fewer pixels. In GPG, transfer success recovered as DFC relaxed the coefficient and surpassed the fixed-coefficient baseline with fewer pixels. DFC showed no consistent gain for EGS-TSSA, which limits perturbations to source-model feature-importance regions.

### Source excerpt

arXiv:2611.25001v1 Announce Type: new Abstract: Generator-based sparse adversarial attacks train encoder--decoder networks to produce transferable sparse perturbations for black-box robustness evaluation. Their perturbation density, however, is set only indirectly by a fixed sparsity coefficient whose mapping to density is unknown and shifts during training, so reaching a target density requires re-tuning; this fixed pressure also yields little transferability gain once density reduction begins. We propose Density Feedback Control (DFC), which tightens or relaxes this coefficient at the end of each epoch according to the gap between the observed hard-mask density and the target, without changing the generator architecture or loss. At a 10\% target density, DFC brings the mean density of three base attacks within one percentage point of the target in a single run and improves the transfer attack success rate (ASR) of two of them on CNN and Transformer targets, including under per-image $\ell_0$ budgets, while perturbing fewer pixels. Tracing GPG's training, we observe that as DFC relaxes the coefficient, the transfer ASR recovers with the density and overtakes the fixed-coefficient baseline with fewer pixels. For EGS-TSSA, which restricts perturbations to source-model feature-importance regions, DFC yields no consistent gain; we analyze how this restriction interacts with density feedback.

## From real to synthetic: PSG-GAN learns sleep apnea signal dynamics for AI training

DevFeed: [From real to synthetic: PSG-GAN learns sleep apnea signal dynamics for AI training](<https://devfeed.tech/articles/from-real-to-synthetic-psg-gan-learns-sleep-apnea-signal-dynamics-for-ai-training-142797.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.24436>)

Author: Riku Huttunen, Samu Kainulainen, Matias Rusanen, Sami Myllymaa, Juha T\\"oyr\\"as, Timo Lepp\\"anen, Henri Korkalainen

Published: 2026-11-19T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [Neural Network](<https://devfeed.tech/topics/neural-network.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-training](<https://devfeed.tech/tags/ai-training.md>), [algorithm](<https://devfeed.tech/tags/algorithm.md>), [applications](<https://devfeed.tech/tags/applications.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [data](<https://devfeed.tech/tags/data.md>), [diagnostics](<https://devfeed.tech/tags/diagnostics.md>), [generative](<https://devfeed.tech/tags/generative.md>), [healthcare](<https://devfeed.tech/tags/healthcare.md>), [osa](<https://devfeed.tech/tags/osa.md>), [physics-med-ph](<https://devfeed.tech/tags/physics-med-ph.md>), [signal](<https://devfeed.tech/tags/signal.md>), [sleep](<https://devfeed.tech/tags/sleep.md>), [synthetic](<https://devfeed.tech/tags/synthetic.md>)

### AI overview

PSG-GAN synthesizes longer segments of multiple polysomnography signals, aiming to capture interactions associated with obstructive sleep apnea that earlier work on short, individual signals overlooked. The generated signals appeared realistic, with EEG and EOG band powers distributed similarly across sleep stages, and mirrored PPG-based heart-rate changes during and after breathing obstructions. An automatic scoring model trained only on automatically scored synthetic signals achieved sleep-staging κ values of 0.71 and 0.63 and respiratory-event-segmentation κ values of 0.57 and 0.60 on test sets of healthy volunteers and OSA patients, respectively. The reported results suggest potential for synthetic data in developing sleep-diagnostic algorithms.

### Source excerpt

arXiv:2611.24436v1 Announce Type: new Abstract: Synthetic polysomnography (PSG) data has great potential in healthcare applications, for example in substituting real data that is sensitive and challenging to collect. Previous research on synthetic PSG data has been largely focused on short segments of individual signals, overlooking the inter-signal dynamics related to sleep disorders. In this study, we developed a generative adversarial network (PSG-GAN) for co-synthesizing longer segments of multiple PSG signals with diverse and complex characteristics typical of obstructive sleep apnea (OSA). The synthetic signals appeared realistic, and the electroencephalography and electrooculography band powers were similarly distributed across sleep stages. Moreover, the PSG-GAN learned to mirror signal interactions such as changes in photoplethysmography-based heart rate during and after breathing obstructions. An automatic scoring model trained solely with automatically scored synthetic signals reached good performance on sleep staging (\k{appa}=0.71 and 0.63) and respiratory event segmentation (\k{appa}=0.57 and 0.60), on two test sets consisting of healthy volunteers and OSA patients, respectively. This demonstrates the potential of synthetic data in supporting algorithm development for sleep diagnostics. The source code is available at https://github.com/rikuhuttunen/psg-gan.

## Quantum Computing-Based Wasserstein Generative Adversarial Networks for Airfoil Parameterization

DevFeed: [Quantum Computing-Based Wasserstein Generative Adversarial Networks for Airfoil Parameterization](<https://devfeed.tech/articles/quantum-computing-based-wasserstein-generative-adversarial-networks-for-airfoil-parameterization-139072.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.20602>)

Author: Abdur Rahman Ovy, Xiaosong Du

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [airfoil](<https://devfeed.tech/tags/airfoil.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [design](<https://devfeed.tech/tags/design.md>), [distribution](<https://devfeed.tech/tags/distribution.md>), [generator](<https://devfeed.tech/tags/generator.md>), [quantum-circuits](<https://devfeed.tech/tags/quantum-circuits.md>), [quantum-computing](<https://devfeed.tech/tags/quantum-computing.md>)

### AI overview

The proposed Q-WGAN uses parameterized quantum circuits for its generator and a classical discriminator to parameterize airfoils represented by free-form deformation control points. On optimal airfoils from 992 aerodynamic design optimizations, it reports 92.1% mean fitting accuracy for control points and 99.2% for airfoil coordinates, close to classical WGAN results of 92.28% and 99.26%, respectively. Q-WGAN also has lower Jensen-Shannon divergence, Fréchet distance, and Wasserstein distance than the classical model, suggesting closer agreement with the reference airfoil distribution. The supplied abstract reports these results but does not provide further methodological details.

### Source excerpt

arXiv:2611.20602v1 Announce Type: new Abstract: Airfoil geometry parameterization is essential in aerodynamic design optimization. Optimization search in the large space suffers from low efficiency and modest feasible region, which could further incur failures to convergence. Generative adversarial networks (GAN), aiming to learn the statistical distribution of existing data, pave a new avenue for shape generation and airfoil parameterization. GAN manages to implicitly reduce design space by generating only smooth, realistic airfoils. Meanwhile, quantum computing has been emergingly exhibiting great potential for achieving superior performance in specific types of engineering problems but has not been introduced to airfoil parameterization. To fill this gap, this work develops quantum computing-based Wasserstein GAN (Q-WGAN) and investigates the performance for airfoil parameterization. The Q-WGAN adopts parameterized quantum circuits as the generator to promote the generative performance, while the discriminator remains classical. The developed Q-WGAN is demonstrated on a dataset of optimal airfoils parameterized by free-form deformation (FFD) control points, resulting from 992 airfoil aerodynamic design optimizations. Results show that Q-WGAN achieves 92.1% mean fitting accuracy on the optimal FFD control points and 99.2% accuracy on airfoil coordinates, comparable with the classical WGAN performance of 92.28% on FFD control points and 99.26% on airfoil coordinates. Q-WGAN achieves lower Jensen-Shannon divergence, Fr\'echet distance, and Wasserstein distance than the classical WGAN, indicating an improved agreement with the reference airfoil distribution. Therefore, this work presents a convincing evidence that Q-WGAN makes a viable option for parameterizing airfoils and possesses a great potential for other relevant engineering applications.

## Well-Log Reconstruction with Pre-trained Geological Knowledge: A Knowledge-Conditioned Large-Language-Model Generative Agent

DevFeed: [Well-Log Reconstruction with Pre-trained Geological Knowledge: A Knowledge-Conditioned Large-Language-Model Generative Agent](<https://devfeed.tech/articles/well-log-reconstruction-with-pre-trained-geological-knowledge-a-knowledge-conditioned-large-language-model-generative-agent-137474.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.19624>)

Author: Yongan Zhang, Yitao Yu, Yizheng Li, An Gong, Youzhuang Sun

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Large Language Model](<https://devfeed.tech/topics/llm.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [LLM evaluation / benchmarking](<https://devfeed.tech/topics/llm-evaluation-benchmarking.md>), [AI Foundation Models](<https://devfeed.tech/topics/ai-foundation-models.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [agent](<https://devfeed.tech/tags/agent.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [conditional](<https://devfeed.tech/tags/conditional.md>), [coupling](<https://devfeed.tech/tags/coupling.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [deep](<https://devfeed.tech/tags/deep.md>), [deep-learning](<https://devfeed.tech/tags/deep-learning.md>), [deep-learning-models](<https://devfeed.tech/tags/deep-learning-models.md>), [domain](<https://devfeed.tech/tags/domain.md>), [feature](<https://devfeed.tech/tags/feature.md>), [foundation](<https://devfeed.tech/tags/foundation.md>), [generative](<https://devfeed.tech/tags/generative.md>), [knowledge](<https://devfeed.tech/tags/knowledge.md>), [language](<https://devfeed.tech/tags/language.md>), [language-model](<https://devfeed.tech/tags/language-model.md>), [large](<https://devfeed.tech/tags/large.md>), [large-language-model](<https://devfeed.tech/tags/large-language-model.md>), [log](<https://devfeed.tech/tags/log.md>), [model](<https://devfeed.tech/tags/model.md>), [well](<https://devfeed.tech/tags/well.md>)

### AI overview

GeoLog-Agent uses geological priors from pre-trained large language models to guide reconstruction of missing well-log intervals. Its Geo-Harness converts stratigraphic reasoning into formation-anchored petrophysical trajectories, feature-wise linear modulation, and conditional adversarial constraints. On a Norwegian North Sea benchmark, it reports mean R² of 0.9307, 0.0920 above the nine-baseline average, with 30.97% lower average RMSE; against the strongest baseline, R² gains are 0.0274-0.0498 and RMSE reductions 16.25%-24.05%. Ablations support contributions from geological reasoning, multi-scale dilated generation, and adversarial modules. Applying the coupling mechanism to six conventional deep-learning models improved R² by 0.0162-0.0789 with cross-well stability. Evidence is limited to the abstract; broader evaluation details are not available here.

### Source excerpt

arXiv:2611.19624v1 Announce Type: cross Abstract: Missing well-log intervals make subsurface reconstruction ill-posed. Existing numerical regression models lack geological prior knowledge to constrain ambiguous inverse mappings. We propose GeoLog-Agent, a knowledge-conditioned generative agent using pre-trained large language models (LLMs) as a source of geological priors. A universal Geo-Harness translates LLM stratigraphic reasoning into formation-anchored petrophysical trajectories, feature-wise linear modulation, and conditional adversarial constraints. On the Norwegian North Sea benchmark, GeoLog-Agent achieves a mean $R^2$ of 0.9307 and a peak $R^2$ of 0.9713. It exceeds the nine-baseline average $R^2$ of 0.8387 by 0.0920, with a 30.97% average RMSE reduction, and exceeds the strongest baseline by 0.0274--0.0498 in $R^2$, with 16.25%--24.05% lower RMSE. Ablations confirm the effectiveness and synergy of geological prior reasoning, multi-scale dilated generation, and conditional adversarial modules. Transferring the coupling mechanism to six conventional deep learning models improves their $R^2$ by 0.0162--0.0789 with cross-well stability. These results demonstrate how foundation-model domain reasoning can constrain deep generative reconstruction of missing petrophysical measurements.

## Lipschitz-Constrained Discriminators in Generative Modeling: Asymptotic Behavior and Support Mismatch

DevFeed: [Lipschitz-Constrained Discriminators in Generative Modeling: Asymptotic Behavior and Support Mismatch](<https://devfeed.tech/articles/lipschitz-constrained-discriminators-in-generative-modeling-asymptotic-behavior-and-support-mismatch-135896.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.18394>)

Author: Eduardo Pinto, Panagiota Birmpa

Published: 2026-11-16T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [behavior](<https://devfeed.tech/tags/behavior.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [distribution](<https://devfeed.tech/tags/distribution.md>), [distributions](<https://devfeed.tech/tags/distributions.md>), [generative](<https://devfeed.tech/tags/generative.md>), [kl-divergence](<https://devfeed.tech/tags/kl-divergence.md>), [metric](<https://devfeed.tech/tags/metric.md>), [model](<https://devfeed.tech/tags/model.md>), [modeling](<https://devfeed.tech/tags/modeling.md>), [models](<https://devfeed.tech/tags/models.md>), [overlap](<https://devfeed.tech/tags/overlap.md>), [stat-ml](<https://devfeed.tech/tags/stat-ml.md>), [strategies](<https://devfeed.tech/tags/strategies.md>), [support](<https://devfeed.tech/tags/support.md>)

### AI overview

The paper characterizes how optimal discriminators for L-Lipschitz-regularized KL divergences behave as the Lipschitz bound L increases, relaxing geometric regularization toward the KL divergence. It considers three regimes: Q absolutely continuous with respect to P and its density bounded above and away from zero; discrete distributions with Q absolutely continuous with respect to P; and discrete distributions without absolute continuity. The results are motivated by adversarial generative modeling, where empirical data distributions are discrete and the model distribution may initially have little or no support overlap. The supplied text does not state the specific asymptotic forms or training strategies.

### Source excerpt

arXiv:2611.18394v1 Announce Type: cross Abstract: We study the asymptotic behavior of optimal discriminators associated to $L-$Lipschitz regularized KL-divergences of two probability distributions $Q$ and $P$ that interpolate between the rescaled Wasserstein$-1$ metric by some parameter $L>0$ and the KL divergence. The parameter $L$ plays the role of a Lipschitz constraint and is viewed as a regularization parameter: small values of $L$ enforce a more geometry-aware discriminator, while as $L$ increases the geometric regularization is progressively relaxed. It is a known result [3] that such divergences converge to the KL divergence as $L\to\infty$. In this paper, we characterize the asymptotic behavior of the optimal discriminators as $L\to\infty$ in different regimes: Firstly, under $Q\ll P$ and the Radon-Nikodym derivative $\frac{dQ}{dP}$ is bounded from above and bounded away from zero. Secondly, when $Q$ and $P$ are discrete distributions satisfying $Q\ll P$, and finally when $Q$ and $P$ are non absolutely continuous discrete distributions. This result motivates training strategies which involve discriminators, such as adversarial generative models. In practice, the data distribution is often represented by a discrete empirical measure, while, especially at the early stages of the training, the model distribution may have little or no overlap with the empirical data distribution.

## Differential Diagnosis of Frontotemporal Dementia and Alzheimer's Disease using Generative Adversarial Network

DevFeed: [Differential Diagnosis of Frontotemporal Dementia and Alzheimer's Disease using Generative Adversarial Network](<https://devfeed.tech/articles/differential-diagnosis-of-frontotemporal-dementia-and-alzheimer-s-disease-using-generative-adversarial-network-128472.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2109.05627>)

Author: Da Ma, Donghuan Lu, Karteek Popuri, Mirza Faisal Beg

Published: 2026-11-11T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [Medical imaging](<https://devfeed.tech/topics/medical-imaging.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [classification](<https://devfeed.tech/tags/classification.md>), [cross-validation](<https://devfeed.tech/tags/cross-validation.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [data-augmentation](<https://devfeed.tech/tags/data-augmentation.md>), [deep-learning](<https://devfeed.tech/tags/deep-learning.md>), [diagnosis](<https://devfeed.tech/tags/diagnosis.md>), [disease](<https://devfeed.tech/tags/disease.md>), [eess-iv](<https://devfeed.tech/tags/eess-iv.md>), [features](<https://devfeed.tech/tags/features.md>), [framework](<https://devfeed.tech/tags/framework.md>), [generative](<https://devfeed.tech/tags/generative.md>), [image](<https://devfeed.tech/tags/image.md>), [imaging](<https://devfeed.tech/tags/imaging.md>), [magnetic](<https://devfeed.tech/tags/magnetic.md>), [medical](<https://devfeed.tech/tags/medical.md>), [multiple](<https://devfeed.tech/tags/multiple.md>), [network](<https://devfeed.tech/tags/network.md>), [performance](<https://devfeed.tech/tags/performance.md>), [q-bio-nc](<https://devfeed.tech/tags/q-bio-nc.md>), [scale](<https://devfeed.tech/tags/scale.md>), [study](<https://devfeed.tech/tags/study.md>), [synthetic](<https://devfeed.tech/tags/synthetic.md>)

### AI overview

The study proposes a GAN-based framework to classify frontotemporal dementia, Alzheimer's disease, and normal controls from MRI. It combines volumetric structural features extracted at multiple scales with synthetic data augmentation. In 10-fold cross-validation on 1,954 images, the framework achieved high accuracy; the authors report that combining multi-scale features and GAN-based augmentation improved performance on dementia subtype differentiation. The excerpt does not specify accuracy values or further validation details.

### Source excerpt

arXiv:2109.05627v2 Announce Type: cross Abstract: Frontotemporal dementia and Alzheimer's disease are two common forms of dementia and are easily misdiagnosed as each other due to their similar pattern of clinical symptoms. Differentiating between the two dementia types is crucial for determining disease-specific intervention and treatment. Recent development of Deep-learning-based approaches in the field of medical image computing are delivering some of the best performance for many binary classification tasks, although its application in differential diagnosis, such as neuroimage-based differentiation for multiple types of dementia, has not been explored. In this study, a novel framework was proposed by using the Generative Adversarial Network technique to distinguish FTD, AD and normal control subjects, using volumetric features extracted at coarse-to-fine structural scales from Magnetic Resonance Imaging scans. Experiments of 10-folds cross-validation on 1,954 images achieved high accuracy. With the proposed framework, we have demonstrated that the combination of multi-scale structural features and synthetic data augmentation based on generative adversarial network can improve the performance of challenging tasks such as differentiating Dementia sub-types.

## Attacking Slide Puzzle CAPTCHAs Based on Time and Space Generative Adversarial Networks

DevFeed: [Attacking Slide Puzzle CAPTCHAs Based on Time and Space Generative Adversarial Networks](<https://devfeed.tech/articles/attacking-slide-puzzle-captchas-based-on-time-and-space-generative-adversarial-networks-126954.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.11355>)

Author: Rui Guo, Xuelian Xie, Botao Xu, Kesheng Liu, Wenxian Wang, Haizhou Wang

Published: 2026-11-10T05:00:00Z

Content type: news

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [Adversarial attacks](<https://devfeed.tech/topics/adversarial-attacks.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [Security](<https://devfeed.tech/topics/security.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [captchas](<https://devfeed.tech/tags/captchas.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [defense](<https://devfeed.tech/tags/defense.md>), [generative](<https://devfeed.tech/tags/generative.md>), [networks](<https://devfeed.tech/tags/networks.md>), [puzzle](<https://devfeed.tech/tags/puzzle.md>), [security](<https://devfeed.tech/tags/security.md>)

### AI overview

The paper reports a method that cracked six named slide puzzle CAPTCHA systems with success rates above 90% and an attack time of about five seconds. It analyzes their defenses, builds separate notch detection and slide trajectory datasets for each system, uses YOLOv11 to locate notches, and trains GANs to generate trajectories intended to learn human sliding characteristics, including motion in two-dimensional space. The abstract also says the authors propose ways to strengthen CAPTCHA security; it does not specify those methods or provide further experimental details.

### Source excerpt

arXiv:2611.11355v1 Announce Type: new Abstract: CAPTCHAs can protect websites from malicious attacks. As a typical representative of behavior CAPTCHAs, slide puzzle CAPTCHAs have become increasingly popular in recent years due to their user-friendliness and high security. Previous attack methods on slide puzzle CAPTCHA mainly face challenges in the following four aspects. (1) The defense mechanisms of previous slide puzzle CAPTCHAs were relatively simple. The current slide puzzle CAPTCHAs have added complex interferences, detected slide trajectory features in the two-dimensional space, and added other defense mechanisms. (2) The previous notch detection methods were difficult to accurately locate the notch positions. (3) Previous attack trajectories were limited to the horizontal dimension, making them impossible to pass the existing detection. (4) The previous attack trajectories were generated by fitting functions, which were relatively simple and could not well simulate the slide characteristics of human users. In response to the above challenges, firstly, we systematically analyze the defense strategies of six most popular slide puzzle CAPTCHAs: Capy, Dingxiang, Geetest, Kyger, Shumei and Tencent. Secondly, for each type of CAPTCHA, we construct a notch detection dataset and a slide trajectory dataset separately. Thirdly, we use advanced YOLOv11 to detect the notch positions and train generative adversarial networks (GAN) to generate attack trajectories. As far as we know, we are the first to generate attack trajectories by directly learning human slide trajectory characteristic. According to the experimental results, this method cracked six most popular slide puzzle CAPTCHAs with success rates of over 90\%, and the attack time is only about 5 seconds, indicating that the security of existing slide puzzle CAPTCHAs is facing challenge. At the same time, we propose various methods to enhance the security of slide puzzle CAPTCHAs.

## Generative Reconstruction of Full-Field Sensor States for Safety-Critical Digital Twins

DevFeed: [Generative Reconstruction of Full-Field Sensor States for Safety-Critical Digital Twins](<https://devfeed.tech/articles/generative-reconstruction-of-full-field-sensor-states-for-safety-critical-digital-twins-123263.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.08499>)

Author: Umme Mahbuba Nabila, Linyu Lin, Majdi I. Radaideh

Published: 2026-11-09T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [VAE](<https://devfeed.tech/topics/vae.md>), [World models](<https://devfeed.tech/topics/world-models.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [autoencoder](<https://devfeed.tech/tags/autoencoder.md>), [conditional](<https://devfeed.tech/tags/conditional.md>), [critical](<https://devfeed.tech/tags/critical.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [cs-sy](<https://devfeed.tech/tags/cs-sy.md>), [diffusion](<https://devfeed.tech/tags/diffusion.md>), [digital](<https://devfeed.tech/tags/digital.md>), [digital-twins](<https://devfeed.tech/tags/digital-twins.md>), [eess-sy](<https://devfeed.tech/tags/eess-sy.md>), [field](<https://devfeed.tech/tags/field.md>), [full](<https://devfeed.tech/tags/full.md>), [generative](<https://devfeed.tech/tags/generative.md>), [generative-models](<https://devfeed.tech/tags/generative-models.md>), [gnn](<https://devfeed.tech/tags/gnn.md>), [graph](<https://devfeed.tech/tags/graph.md>), [modeling](<https://devfeed.tech/tags/modeling.md>), [models](<https://devfeed.tech/tags/models.md>), [network](<https://devfeed.tech/tags/network.md>), [neural](<https://devfeed.tech/tags/neural.md>), [neural-network](<https://devfeed.tech/tags/neural-network.md>), [safety](<https://devfeed.tech/tags/safety.md>), [sensor](<https://devfeed.tech/tags/sensor.md>)

### AI overview

The study evaluates six methods for reconstructing missing temperature measurements at Oregon State University's High Temperature Test Facility, both within one experiment and across experiments. For reconstruction within the PG-29 depressurized cooldown test, a graph neural network-GRU variational autoencoder performs best (MAE 11.73°C, RMSE 20.54°C, R² 0.98). The graph-based deterministic model also outperforms a plain GRU, while all three generative models reduce RMSE by more than half versus ridge regression. A diffusion model retains R² 0.94 when available sensors fall from 91 to 45. For transfer from PG-27 pressurized cooldown to PG-29, a diffusion model conditioned on section context achieves the strongest reported result (RMSE 42.84°C, R² 0.91). The evidence is limited to the abstract and these reported test protocols.

### Source excerpt

arXiv:2611.08499v1 Announce Type: new Abstract: Accurate reconstruction of incomplete sensor fields is critical for state awareness in safety-critical digital twins (DTs). Thermocouples in high-temperature and radiation environments may drift, fail, or become intermittently unavailable, causing partial observability. This study develops a sparse-sensor reconstruction framework for Oregon State University's High Temperature Test Facility (HTTF) and evaluates six methods under intra- and inter-experiment protocols: three conditional generative models (i.e., a diffusion model [DM], variational autoencoder [VAE], and generative adversarial network [GAN]), two deterministic recurrent neural-network models, and ridge regression. For intra-experiment reconstruction within the PG{-}29 depressurized conduction cooldown (DCC) test, the models combine graph-based spatial encoding with gated recurrent units. The graph neural network (GNN)--gated recurrent unit (GRU) VAE achieves the lowest error, with a mean absolute error (MAE) of $11.73,^\circ$C, a root mean square error (RMSE) of $20.54,^\circ$C, and an $R^2$ of 0.98. The deterministic GNN--GRU and plain GRU obtain RMSE values of $25.26,^\circ$C and $35.22,^\circ$C, respectively, demonstrating the benefit of spatial modeling. All three generative models reduce RMSE by more than half compared with ridge regression. The DM maintains an $R^2$ of 0.94 when known sensors decrease from 91 to 45. For inter-experiment reconstruction, models trained on PG{-}27 pressurized conduction cooldown (PCC) are evaluated on PG{-}29 using section-context conditioning. The section-context DM achieves the best transfer performance, with an RMSE of $42.84,^\circ$C and an $R^2$ of 0.91. These results demonstrate the value of spatiotemporal generative modeling for missing-sensor reconstruction in safety-critical DT applications.

## Continuous Adversarial MeanFlow for Few-Step Generative Adaptation from Any Parameterization

DevFeed: [Continuous Adversarial MeanFlow for Few-Step Generative Adaptation from Any Parameterization](<https://devfeed.tech/articles/continuous-adversarial-meanflow-for-few-step-generative-adaptation-from-any-parameterization-120474.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2608.19540>)

Author: Yara Bahram, Zahra Dehghani, M\\'elodie Desbos, Eric Granger, Pablo Piantanida, Mohammadhadi Shateri

Published: 2026-11-06T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [Neural Network](<https://devfeed.tech/topics/neural-network.md>), [ai-image-generator](<https://devfeed.tech/topics/ai-image-generator.md>)

Tags: [about](<https://devfeed.tech/tags/about.md>), [abstract](<https://devfeed.tech/tags/abstract.md>), [acceleration](<https://devfeed.tech/tags/acceleration.md>), [adapting](<https://devfeed.tech/tags/adapting.md>), [adversarial](<https://devfeed.tech/tags/adversarial.md>), [also](<https://devfeed.tech/tags/also.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [based](<https://devfeed.tech/tags/based.md>), [change](<https://devfeed.tech/tags/change.md>), [compare](<https://devfeed.tech/tags/compare.md>), [continuous](<https://devfeed.tech/tags/continuous.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [data](<https://devfeed.tech/tags/data.md>), [diffusion](<https://devfeed.tech/tags/diffusion.md>), [domain](<https://devfeed.tech/tags/domain.md>), [domain-model](<https://devfeed.tech/tags/domain-model.md>), [domains](<https://devfeed.tech/tags/domains.md>), [evaluations](<https://devfeed.tech/tags/evaluations.md>), [existing](<https://devfeed.tech/tags/existing.md>), [fast](<https://devfeed.tech/tags/fast.md>), [generative](<https://devfeed.tech/tags/generative.md>)

### AI overview

MeanFlow-Transfer (MF-T) adapts pretrained diffusion or flow models across parameterizations through a common velocity map while compressing them into few-step MeanFlow generators. Continuous Adversarial MeanFlow (CAMF) scores changes in a learned potential over each predicted interval jump, addressing the limits of endpoint criteria on small jumps and instantaneous-velocity criteria on large jumps. It requires no gradient penalty or Jacobian-vector product, is reported to remain well-conditioned across interval lengths, and recovers the instantaneous criterion as intervals shrink. Across four ImageNet-based source models and five target domains, MF-T + CAMF approached or exceeded fine-tuned many-step teachers on FID and FDD with up to 125 times fewer Neural Function Evaluations; CAMF also improved MF-T few-step FID by about 30% on average and outperformed prior adversarial post-training on iMF at every evaluated NFE.

### Source excerpt

arXiv:2608.19540v4 Announce Type: replace Abstract: Training fast generators on new domains with limited data remains challenging for two reasons. First, adapting a pretrained diffusion or flow model leaves its costly multi-step sampling intact, and existing acceleration methods are tied to the source parameterization (x, epsilon, v, or u), so heterogeneous pretrained models share no common acceleration target. Second, while adversarial post-training sharpens generation quality, existing criteria each cover only one end of the step range. Discrete criteria compare jump endpoints and lose their signal as the jump shrinks, while continuous criteria score the instantaneous velocity and misjudge the large jumps of few-step sampling. Neither fits MeanFlow (MF) models, which predict average velocities over intervals of any length. We propose MeanFlow-Transfer (MF-T), which adapts a source domain model of any parameterization and compresses it into a few-step MF generator through a common velocity map. We then introduce Continuous Adversarial MeanFlow (CAMF), which scores the change of a learned potential along each predicted interval jump of MF. CAMF needs no gradient penalty or Jacobian-vector product (JVP). We also prove it stays well-conditioned for every interval length and recovers the instantaneous criterion as the interval vanishes. Adapting four ImageNet-based source models (DiT, SiT, JiT, iMF) to five target domains, MF-T + CAMF approaches or exceeds the fine-tuned many-step teacher in FID and FDD at up to 125x fewer Neural Function Evaluations (NFEs). Further, CAMF improves MF-T's few-step FID by about 30% on average, outperforming prior adversarial post-training on iMF at every NFE.

## Bridging the Simulation-to-Experiment Gap with Generative Models using Adversarial Distribution Alignment

DevFeed: [Bridging the Simulation-to-Experiment Gap with Generative Models using Adversarial Distribution Alignment](<https://devfeed.tech/articles/bridging-the-simulation-to-experiment-gap-with-generative-models-using-adversarial-distribution-alignment-108485.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2604.01169>)

Author: Kai Nelson, Tobias Kreiman, Sergey Levine, Aditi S. Krishnapriyan

Published: 2026-10-30T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [AI Models](<https://devfeed.tech/topics/ai-models.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [Simulation](<https://devfeed.tech/topics/simulation.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [alignment](<https://devfeed.tech/tags/alignment.md>), [approach](<https://devfeed.tech/tags/approach.md>), [atomic](<https://devfeed.tech/tags/atomic.md>), [bridging](<https://devfeed.tech/tags/bridging.md>), [challenge](<https://devfeed.tech/tags/challenge.md>), [cond-mat-mtrl-sci](<https://devfeed.tech/tags/cond-mat-mtrl-sci.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [data](<https://devfeed.tech/tags/data.md>), [data-driven](<https://devfeed.tech/tags/data-driven.md>), [distribution](<https://devfeed.tech/tags/distribution.md>), [domain](<https://devfeed.tech/tags/domain.md>), [engineering](<https://devfeed.tech/tags/engineering.md>), [experiment](<https://devfeed.tech/tags/experiment.md>), [experimental](<https://devfeed.tech/tags/experimental.md>), [framework](<https://devfeed.tech/tags/framework.md>), [gap](<https://devfeed.tech/tags/gap.md>), [generative](<https://devfeed.tech/tags/generative.md>), [generative-models](<https://devfeed.tech/tags/generative-models.md>), [measurements](<https://devfeed.tech/tags/measurements.md>), [models](<https://devfeed.tech/tags/models.md>), [q-bio-bm](<https://devfeed.tech/tags/q-bio-bm.md>), [simulation](<https://devfeed.tech/tags/simulation.md>)

### AI overview

Adversarial Distribution Alignment (ADA) addresses the simulation-to-experiment gap by pretraining a generative model on fully observed but imperfect simulation data, then aligning its distribution with partial observations from real experiments. In the physical-sciences application, the model starts from atomic positions sampled from a simulated Boltzmann distribution. The authors prove recovery of the target observable distribution, including when observables are multiple and potentially correlated, and report empirical validation on synthetic, molecular, and experimental protein data. The excerpt does not provide implementation details or quantify performance.

### Source excerpt

arXiv:2604.01169v2 Announce Type: replace Abstract: A fundamental challenge in science and engineering is the simulation-to-experiment gap. While we often possess prior knowledge of physical laws, these physical laws can be too difficult to solve exactly for complex systems. Such systems are commonly modeled using simulators, which impose computational approximations. Meanwhile, experimental measurements more faithfully represent the real world, but experimental data typically consists of observations that only partially reflect the system's full underlying state. We propose a data-driven distribution alignment framework that bridges this simulation-to-experiment gap by pre-training a generative model on fully observed (but imperfect) simulation data, then aligning it with partial (but real) observations of experimental data. While our method is domain-agnostic, we ground our approach in the physical sciences by introducing Adversarial Distribution Alignment (ADA). This method aligns a generative model of atomic positions -- initially trained on a simulated Boltzmann distribution -- with the distribution of experimental observations. We prove that our method recovers the target observable distribution, even with multiple, potentially correlated observables. We also empirically validate our framework on synthetic, molecular, and experimental protein data, demonstrating that it can align generative models with diverse observables. Our code is available at https://kaityrusnelson.com/ada/.

## Generalization in Learning: Eyring Formula and Predator-prey Model

DevFeed: [Generalization in Learning: Eyring Formula and Predator-prey Model](<https://devfeed.tech/articles/generalization-in-learning-eyring-formula-and-predator-prey-model-103894.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.35965>)

Author: S. V. Kozyrev, I. A. Lopatin, Alexander Pechen

Published: 2026-10-28T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [generalization in machine learning](<https://devfeed.tech/topics/generalization-in-machine-learning.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>)

Tags: [approximation](<https://devfeed.tech/tags/approximation.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [generalization-in-machine-learning](<https://devfeed.tech/tags/generalization-in-machine-learning.md>), [generative](<https://devfeed.tech/tags/generative.md>), [gradient-descent](<https://devfeed.tech/tags/gradient-descent.md>), [grokking](<https://devfeed.tech/tags/grokking.md>), [model](<https://devfeed.tech/tags/model.md>)

### AI overview

The work investigates machine learning generalization through a predator-prey model. For one example--recovering an unknown function with a fixed-degree polynomial from noisy data--the model yields a better approximation than gradient descent. The authors also give biological and chemical interpretations of stochastic gradient Langevin dynamics and generative adversarial networks, and use thermodynamic arguments, especially the Eyring formula, to explain grokking (delayed generalization). These findings are reported for the described example; the supplied abstract does not establish broader comparative performance.

### Source excerpt

arXiv:2610.35965v1 Announce Type: new Abstract: In this work, generalization in machine learning is investigated using a predator-prey model. We overview biological and chemical interpretation of the stochastic gradient Langevin dynamics and generative adversarial network. For a particular example of recovering an unknown function by a polynomial of fixed degree from a set of noisy data, the predator-prey model provides better approximation compared to the gradient descent method. Further, thermodynamical arguments, in particular Eyring formula, are used to explain grokking (delayed generalization) phenomenon in machine learning.

## MRI-to-CT Translation of Spinal Images via Anatomy-Guided Pseudo-Pairing and Adversarial Diffusion

DevFeed: [MRI-to-CT Translation of Spinal Images via Anatomy-Guided Pseudo-Pairing and Adversarial Diffusion](<https://devfeed.tech/articles/mri-to-ct-translation-of-spinal-images-via-anatomy-guided-pseudo-pairing-and-adversarial-diffusion-103235.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.37134>)

Author: Eduardo Nieto, Ibai Genua, Davide Scorza, Karen L\\'opez-Linares

Published: 2026-10-28T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Medical imaging](<https://devfeed.tech/topics/medical-imaging.md>), [Image processing](<https://devfeed.tech/topics/image-processing.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [3D visualization](<https://devfeed.tech/topics/3d-visualization.md>)

Tags: [3d](<https://devfeed.tech/tags/3d.md>), [adversarial](<https://devfeed.tech/tags/adversarial.md>), [alignment](<https://devfeed.tech/tags/alignment.md>), [coherence](<https://devfeed.tech/tags/coherence.md>), [consistency](<https://devfeed.tech/tags/consistency.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [data](<https://devfeed.tech/tags/data.md>), [diffusion](<https://devfeed.tech/tags/diffusion.md>), [distribution](<https://devfeed.tech/tags/distribution.md>), [framework](<https://devfeed.tech/tags/framework.md>), [images](<https://devfeed.tech/tags/images.md>), [imaging](<https://devfeed.tech/tags/imaging.md>), [matching](<https://devfeed.tech/tags/matching.md>), [metrics](<https://devfeed.tech/tags/metrics.md>), [model](<https://devfeed.tech/tags/model.md>), [network](<https://devfeed.tech/tags/network.md>), [pairing](<https://devfeed.tech/tags/pairing.md>), [pairs](<https://devfeed.tech/tags/pairs.md>), [planning](<https://devfeed.tech/tags/planning.md>), [post](<https://devfeed.tech/tags/post.md>), [translation](<https://devfeed.tech/tags/translation.md>)

### AI overview

The framework synthesizes CT-like spinal images from MRI using three stages: anatomy-guided matching and rigid registration to create pseudo-pairs from unpaired scans, adversarial diffusion for translation, and 3D refinement to improve continuity across slices. It was trained on over 10,000 unpaired sagittal slices and evaluated on 1,911 slices from 83 MRI volumes across spinal regions. Compared with diffusion alone, it raised binary 3D vertebral Dice from 0.666 to 0.739; the authors also report fewer visible slice discontinuities and improved vertebral coherence in axial and coronal views. The abstract presents this as support for further development of MRI-derived CT-like visualization for preoperative planning, while noting the need to preserve fine intensity detail alongside volumetric coherence.

### Source excerpt

arXiv:2610.37134v1 Announce Type: cross Abstract: Magnetic Resonance Imaging (MRI) is routinely used to diagnose spinal disorders, but surgical planning often requires Computed Tomography (CT) for better visualization of vertebral bone anatomy. Synthesizing CT from MRI could avoid additional radiation exposure and streamline preoperative assessment. We propose a three-stage framework for spine MRI-to-CT synthesis. First, an anatomy-guided vertebral matching system identifies similar sagittal MRI and CT slices using spine segmentations and aligns them through rigid registration, creating pseudo-paired data from unpaired scans. Second, these pairs train an adversarial diffusion model for MRI-to-CT translation. Third, synthesized slices are assembled into 3D volumes and refined by a network that improves inter-slice continuity and structural coherence. The framework was trained on over 10,000 unpaired sagittal MRI and CT slices and evaluated on 1,911 slices from 83 MRI volumes spanning all spinal regions. Assessment combined distribution-based measures with vertebral segmentation metrics. Compared with diffusion alone, the proposed approach increased binary 3D vertebral Dice from 0.666 to 0.739. Volumetric post-processing reduced visible discontinuities between stacked slices and improved vertebral coherence in axial and coronal views. Combining anatomy-guided pseudo-pairing, adversarial diffusion and 3D refinement improves spine MRI-to-CT synthesis. Vertebral matching primarily enhanced anatomical alignment and fidelity, while volumetric refinement improved 3D consistency. These findings support further development of MRI-derived CT-like visualization for preoperative planning, particularly to preserve fine intensity details while maintaining volumetric coherence.

## Spatial disparities in simulated patrol detection: a reproducible generative-model audit across Baltimore and Chicago

DevFeed: [Spatial disparities in simulated patrol detection: a reproducible generative-model audit across Baltimore and Chicago](<https://devfeed.tech/articles/spatial-disparities-in-simulated-patrol-detection-a-reproducible-generative-model-audit-across-baltimore-and-chicago-99780.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2603.18987>)

Author: Pronob Kumar Barman, Pronoy Kumar Barman

Published: 2026-10-26T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [geospatial](<https://devfeed.tech/topics/geospatial.md>), [audit](<https://devfeed.tech/topics/audit.md>), [Synthetic Data Generation](<https://devfeed.tech/topics/synthetic-data-generation.md>)

Tags: [audit](<https://devfeed.tech/tags/audit.md>), [auditing](<https://devfeed.tech/tags/auditing.md>), [crime](<https://devfeed.tech/tags/crime.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [detection](<https://devfeed.tech/tags/detection.md>), [fairness](<https://devfeed.tech/tags/fairness.md>), [gan](<https://devfeed.tech/tags/gan.md>), [generative](<https://devfeed.tech/tags/generative.md>), [model](<https://devfeed.tech/tags/model.md>), [police](<https://devfeed.tech/tags/police.md>), [probabilistic](<https://devfeed.tech/tags/probabilistic.md>), [simulation](<https://devfeed.tech/tags/simulation.md>), [spatial](<https://devfeed.tech/tags/spatial.md>), [uncertainty](<https://devfeed.tech/tags/uncertainty.md>)

### AI overview

The audit compares simulated patrol allocation and detection across Baltimore and Chicago using a GAN, empirical-location sampling, and uniform-area deployment. Across 560 trajectories, detection-only training produced Black-majority versus White-majority area detection-rate differences from -2.03 to 0.75 percentage points, with total coverage from 1.78% to 7.50%. A conditional tabular model for balanced synthetic replacement failed geographic consistency checks in 107 of 220 seed-month fits, preventing complete annual trajectories. Results varied with observation process, allocation policy, training budget, and random seed. The simulations enable reproducible scenario comparisons but do not estimate individual racial disparities in police contact or the distribution of unrecorded crime; the authors recommend reporting absolute differences, coverage, synthesis feasibility, and simulation uncertainty alongside ratio-based fairness metrics.

### Source excerpt

arXiv:2603.18987v3 Announce Type: replace Abstract: Auditing predictive policing requires distinguishing recorded crime incidents, simulated patrol observations, and the demographic characteristics of places. We present a reproducible spatial simulation audit combining a generative adversarial network (GAN) for patrol allocation with a probabilistic detection model. Using frozen archives of 145,178 Baltimore records (2017-2019) and 234,882 Chicago records (2022), we rebuild geographic covariates from Census tract boundaries and American Community Survey data. We compare GAN allocation with empirical-location sampling and uniform-area deployment under alternative observation streams. Twenty random seeds per condition yield 560 trajectories, each covering February through December. Under detection-only training, mean detection-rate differences between Black-majority and White-majority areas range from -2.03 to 0.75 percentage points across city-years, while total coverage ranges from 1.78% to 7.50%. Paired comparisons with full-record reference trajectories assess observation feedback conditional on the recorded incident archives. In an experiment using a conditional tabular generative model for balanced synthetic replacement, geographic consistency requirements fail in 107 of 220 seed-month fits, leaving no complete annual synthetic-replacement trajectories. Simulated disparities vary with the observation process, allocation policy, training budget, and random seed. These results support reproducible comparisons of modeled scenarios but do not estimate individual racial disparities in police contact or the distribution of unrecorded crime. We recommend reporting absolute detection differences, coverage, synthesis feasibility, and simulation uncertainty alongside ratio-based fairness metrics.

## Revisiting GAN with Bayes-Optimal Discrimination

DevFeed: [Revisiting GAN with Bayes-Optimal Discrimination](<https://devfeed.tech/articles/revisiting-gan-with-bayes-optimal-discrimination-96803.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2510.25609>)

Author: Mohammadreza Tavasoli Naeini, Ali Bereyhi, Morteza Noshad, Ben Liang, Alfred O. Hero III

Published: 2026-10-23T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [architectures](<https://devfeed.tech/tags/architectures.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [bolt](<https://devfeed.tech/tags/bolt.md>), [budget](<https://devfeed.tech/tags/budget.md>), [cifar-10](<https://devfeed.tech/tags/cifar-10.md>), [class](<https://devfeed.tech/tags/class.md>), [classification](<https://devfeed.tech/tags/classification.md>), [coverage](<https://devfeed.tech/tags/coverage.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [data](<https://devfeed.tech/tags/data.md>), [discrimination](<https://devfeed.tech/tags/discrimination.md>), [distributions](<https://devfeed.tech/tags/distributions.md>), [eess-sp](<https://devfeed.tech/tags/eess-sp.md>), [entropy](<https://devfeed.tech/tags/entropy.md>), [error](<https://devfeed.tech/tags/error.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [fake](<https://devfeed.tech/tags/fake.md>), [game](<https://devfeed.tech/tags/game.md>), [gan-training](<https://devfeed.tech/tags/gan-training.md>), [generation](<https://devfeed.tech/tags/generation.md>), [generative](<https://devfeed.tech/tags/generative.md>)

### AI overview

BOLT-GAN replaces the GAN discriminator's cross-entropy objective with a loss that bounds Bayes error: the discriminator tightens the bound while the generator maximizes it. With balanced class priors, the unrestricted population game is equivalent to minimizing total variation; restricting the discriminator to be 1-Lipschitz gives BOLT-GAN, with its discrepancy bounded in relation to Wasserstein-1 on bounded domains. In controlled tests on four image-generation benchmarks, it reportedly improved sample quality and coverage over WGAN-GP under matched architectures and training settings. At 20 epochs, FID fell from 60.0 to 44.2 on CIFAR-10 and from 102.5 to 28.4 on LSUN Bedroom-64.

### Source excerpt

arXiv:2510.25609v4 Announce Type: replace Abstract: We approach generative adversarial network (GAN) training from its original real-versus-fake classification viewpoint, replacing the cross-entropy discriminator objective with an upper bound on the Bayes error rate (BER). Using the Bayes optimal learning threshold (BOLT) loss, we formulate an adversarial game in which the discriminator tightens this bound and the generator maximizes it. We characterize the unrestricted optimal discriminator and show that, under balanced class priors, the population game is equivalent to minimizing total variation between the data and generator distributions. Restricting the bounded discriminator to be 1-Lipschitz yields BOLT-GAN. For balanced priors and the full discriminator class, we establish upper and lower bounds relating its population discrepancy to Wasserstein-1 distance on bounded domains. Controlled experiments on four image-generation benchmarks show improved sample quality and coverage over WGAN-GP under matched architectures, optimization settings, and gradient penalties. At a common 20-epoch budget, BOLT-GAN reduces Fr\'echet Inception Distance (FID) from 60.0 to 44.2 on CIFAR-10 and from 102.5 to 28.4 on LSUN Bedroom-64.

## NeuroCodec: EEG-Guided Latent-Domain Generative Model for High-Fidelity Target Speaker Extraction

DevFeed: [NeuroCodec: EEG-Guided Latent-Domain Generative Model for High-Fidelity Target Speaker Extraction](<https://devfeed.tech/articles/neurocodec-eeg-guided-latent-domain-generative-model-for-high-fidelity-target-speaker-extraction-94986.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.29815>)

Author: Julian Silva, Jaliya Kodithuwakku, Navindu Sirimanna, Shaveen Herath, Wageesha N. Manamperi, Joshua P. Kulasingham

Published: 2026-10-22T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Transformer](<https://devfeed.tech/topics/transformer.md>), [generative-ai-tools](<https://devfeed.tech/topics/generative-ai-tools.md>), [Mamba](<https://devfeed.tech/topics/mamba.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [fourier transform](<https://devfeed.tech/topics/fourier-transform.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [attention](<https://devfeed.tech/tags/attention.md>), [audio](<https://devfeed.tech/tags/audio.md>), [codec](<https://devfeed.tech/tags/codec.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [eeg](<https://devfeed.tech/tags/eeg.md>), [eess-as](<https://devfeed.tech/tags/eess-as.md>), [generative](<https://devfeed.tech/tags/generative.md>), [mamba](<https://devfeed.tech/tags/mamba.md>), [masking](<https://devfeed.tech/tags/masking.md>), [metrics](<https://devfeed.tech/tags/metrics.md>), [neural](<https://devfeed.tech/tags/neural.md>), [objectives](<https://devfeed.tech/tags/objectives.md>), [party](<https://devfeed.tech/tags/party.md>), [processing](<https://devfeed.tech/tags/processing.md>), [real-time](<https://devfeed.tech/tags/real-time.md>), [regions](<https://devfeed.tech/tags/regions.md>), [space](<https://devfeed.tech/tags/space.md>), [spatial](<https://devfeed.tech/tags/spatial.md>), [speaker](<https://devfeed.tech/tags/speaker.md>), [target](<https://devfeed.tech/tags/target.md>)

### AI overview

NeuroCodec uses EEG to guide target speaker extraction by generating speech in a neural audio codec's latent space, addressing fidelity limits of time-frequency masking in overlapping speech and narrowband audio. It models spatial EEG dependencies with graph convolutions and combines EEG and speech representations using attention and Mamba-based temporal modeling. A six-phase curriculum adds latent reconstruction, multi-scale spectral, and adversarial objectives. On the Cocktail Party and KUL datasets, it outperformed baselines on several perceptual metrics; in a 20-person listening study, its mean score was 82.24 versus 49.63 for M3ANet. The abstract reports 44.1 kHz reconstruction with spectral content up to 20 kHz and a causal streaming formulation for real-time processing.

### Source excerpt

arXiv:2610.29815v1 Announce Type: cross Abstract: Driven by recent advancements in hearing assistive devices, target speaker extraction (TSE) in multi-talker environments has drawn significant attention. To address this, hearing technologies identify attention using neural signals recorded via electroencephalography (EEG) to guide TSE, enabling neuro-steered auditory attention decoding. However, most existing techniques rely on time-frequency masking, which suffers from low fidelity in spectro-temporally overlapping regions and is often restricted to narrowband audio. In contrast, we introduce NeuroCodec, which, to the best of our knowledge, is the first generative EEG-guided TSE framework operating in the latent space of a neural audio codec. Specifically, NeuroCodec models spatial EEG dependencies via graph convolutions and fuses cross-modal EEG and speech representations through attention mechanisms and Mamba-based temporal modeling. We further introduce a six-phase training curriculum that progressively incorporates latent reconstruction, multi-scale spectral, and adversarial objectives to ensure stable, high-fidelity speech generation. Experiments on the public Cocktail Party and KUL datasets show NeuroCodec outperforms baselines, across several perceptual metrics. A listening experiment with 20 participants validated that NeuroCodec offers greater perceptual accuracy, achieving a mean score of 82.24 compared with 49.63 for M3ANet. Additionally, the proposed method is seen to enable 44.1 kHz reconstruction with spectral content extending up to 20 kHz, further supporting the perceptual results. Finally, we extend NeuroCodec with a causal streaming formulation suited for real-time processing, providing a basis for hearing assistive devices.

## Adversarial Post-Training of Foundation Models for High-Frequency Trading

DevFeed: [Adversarial Post-Training of Foundation Models for High-Frequency Trading](<https://devfeed.tech/articles/adversarial-post-training-of-foundation-models-for-high-frequency-trading-93511.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.25604>)

Author: Alexandre Bismuth, Kang Li, Aramis Marti-Shahandeh, Valentin Mohl, Jakob Foerster, Anisoara Calinescu

Published: 2026-10-21T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [post-training](<https://devfeed.tech/topics/post-training.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [lora](<https://devfeed.tech/topics/lora.md>), [AI Foundation Models](<https://devfeed.tech/topics/ai-foundation-models.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [foundation-models](<https://devfeed.tech/tags/foundation-models.md>), [high-frequency-trading](<https://devfeed.tech/tags/high-frequency-trading.md>), [limit-order-book](<https://devfeed.tech/tags/limit-order-book.md>), [post-training](<https://devfeed.tech/tags/post-training.md>)

### AI overview

The proposed post-training method uses a Wasserstein GAN critic trained on rollouts through the true, non-differentiable matching engine to guide rank-4 LoRA evolution-strategy updates to a pretrained LOB model, constrained by a KL trust region. Its memory-efficient EGGROLL variant improves held-out LOB-Bench realism by 17.6% (WS-21) and 17.0% (Mean-L1); selection-matched random perturbations and compute-matched continued pretraining show no improvement. Over 500-message rollouts, per-position divergence indicates slower error accumulation, from 3.1x to 2.3x. The abstract reports under 50M training tokens and 12 GPU-hours per seed, compared with a 48B-token pretraining budget. It says ablations and reward-hacking analysis provide design guidance, but the supplied excerpt does not detail those findings.

### Source excerpt

arXiv:2610.25604v1 Announce Type: new Abstract: While autoregressive foundation models trained on limit order book (LOB) data show state-of-the-art performance for market simulation, pre-training with cross-entropy neither penalizes the compounding of sampling errors over free-running rollouts nor targets desired market properties. We introduce a post-training method for LOB foundation models using a Wasserstein generative adversarial network (GAN) objective. A critic, trained online on rollouts through the true, non-differentiable matching engine, guides updates from evolution strategies over rank-4 LoRA perturbations of the pre-trained generator, under a KL trust region. For efficiency, we use EGGROLL (Evolution Guided GeneRal Optimisation via Low-rank Learning), a memory-efficient variant of evolution strategies. On a held-out panel, adversarial post-training improves LOB-Bench realism by 17.6% (WS-21) and 17.0% (Mean-$L_1$), while selection-matched random-perturbation and compute-matched continued-pre-training controls yield no improvement. Post-training is thus a necessary stage that more pre-training does not replace. Per-position divergence measurements show that it slows error accumulation over 500-message rollouts from $3.1\times$ to $2.3\times$. Adversarial post-training is also inexpensive, requiring under 50M training tokens and 12 GPU-hours per seed, three orders of magnitude below the 48B-token pre-training budget. Ablations over critic inputs, optimizers, and the population-rollout trade-off, together with an analysis of reward hacking, yield practical design guidelines for adversarial post-training.

## Marginal Calibration Masks Generator Differences in Held-Out Evaluation of Multivariate Time Series, with an Application to ERA5 Weather Sequences

DevFeed: [Marginal Calibration Masks Generator Differences in Held-Out Evaluation of Multivariate Time Series, with an Application to ERA5 Weather Sequences](<https://devfeed.tech/articles/marginal-calibration-masks-generator-differences-in-held-out-evaluation-of-multivariate-time-series-with-an-application-to-era5-weather-sequences-94167.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2609.33761>)

Author: Mustafa \\"Ozayta\\c{c}, \\"Ozge Karada\\u{g} Ata\\c{s}

Published: 2026-10-21T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Time Series](<https://devfeed.tech/topics/time-series.md>), [Temporal data](<https://devfeed.tech/topics/temporal-data.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [benchmark overfitting machine learning](<https://devfeed.tech/topics/benchmark-overfitting-machine-learning.md>), [Sequences](<https://devfeed.tech/topics/sequences.md>)

Tags: [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [generative-models](<https://devfeed.tech/tags/generative-models.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [time-series](<https://devfeed.tech/tags/time-series.md>), [weather](<https://devfeed.tech/tags/weather.md>)

### AI overview

On held-out ERA5 weather sequences, marginal calibration made KS scores nearly equal to a real-versus-real noise-and-shift floor, leaving neither RCGAN nor TimeGAN distinguishable from StatD2GAN on that metric. The pre-specified evaluation used the final two years with a 168-hour embargo, training-only calibration, and 25 matched location-seed pairs across five locations. Removing StatD2GAN's sorted-representation discriminator significantly worsened cross-variable dependence; re-implemented TimeGAN performed better on Kendall tau and autocorrelation MAE, while RCGAN performed worse on Kendall tau MAE and constraint violations. Calibration increased physical constraint violations; projection removed them but exceeded the stated marginal-change tolerance in three locations. The authors recommend floor-referenced marginal evaluation alongside matched-pair tests and dependence and sequence-level metrics.

### Source excerpt

arXiv:2609.33761v2 Announce Type: replace Abstract: Generative models for multivariate time series are routinely evaluated with pooled distributional metrics computed after marginal calibration. We show that, under a held-out protocol, the calibrated Kolmogorov-Smirnov (KS) distance differs from a real-versus-real floor by at most the distance of the calibrated sample to the training block, so that it carries little information about the generator. We audit a three-discriminator GAN with evolutionary weight adaptation (StatD2GAN) against re-implemented RCGAN and TimeGAN baselines on ERA5 weather sequences under a pre-specified protocol: the final two calendar years are held out behind a 168 hour embargo, calibration is fitted on the training block only, and the decision rules were frozen in advance. Evidence comes from 25 matched (location, seed) pairs across five locations, tested with Holm-corrected Wilcoxon signed-rank tests. First, isotonic calibration drives KS to a median of 1.019 times a per-location noise-and-shift floor, and neither baseline is distinguishable from StatD2GAN on it. Second, the sorted-representation discriminator is the only component whose removal significantly degrades cross-variable dependence (Kendall tau MAE +0.116, Holm p = 0.002); a rank-transformed variant also degrades it, consistent with quantile supervision rather than copula matching. Third, calibration raises physical constraint violations; projection removes them, but the largest marginal change (0.0110, Lhasa) exceeds the pre-specified tolerance in three locations. Fourth, the re-implemented TimeGAN is significantly better than StatD2GAN on Kendall tau and ACF MAE, RCGAN is worse on Kendall tau MAE and constraint violations, and the pre-specified collapse criterion was not met. We recommend floor-referenced marginal evaluation, matched-pair testing, and dependence and sequence-level metrics.

## IonQ | IonQ and Leading Global Automotive Manufacturer Collaborate to Advance Materials Science and Vehicle Durability Using Quantum Generative AI

DevFeed: [IonQ | IonQ and Leading Global Automotive Manufacturer Collaborate to Advance Materials Science and Vehicle Durability Using Quantum Generative AI](<https://devfeed.tech/articles/ionq-ionq-and-leading-global-automotive-manufacturer-collaborate-to-advance-materials-science-and-vehicle-durability-using-quantum-generative-ai-83805.md>)

Original publisher: [Read original article](<https://ionq.com/blog/ionq-and-leading-global-automotive-manufacturer-collaborate-to-advance>)

Author: IonQ Staff

Published: 2026-01-15T23:19:20Z

Content type: article

Language: en

Sources: [IonQ](<https://devfeed.tech/sources/ionq.md>)

Topics: [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>), [Computational Chemistry / Materials Science](<https://devfeed.tech/topics/computational-chemistry-materials-science.md>), [Image processing](<https://devfeed.tech/topics/image-processing.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>), [AI Engineering](<https://devfeed.tech/topics/ai-engineering.md>), [Computer science](<https://devfeed.tech/topics/computer-science.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [ai-model](<https://devfeed.tech/tags/ai-model.md>), [ai-model-training](<https://devfeed.tech/tags/ai-model-training.md>), [ai-models](<https://devfeed.tech/tags/ai-models.md>), [algorithm](<https://devfeed.tech/tags/algorithm.md>), [algorithms](<https://devfeed.tech/tags/algorithms.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [astronomy](<https://devfeed.tech/tags/astronomy.md>), [automotive](<https://devfeed.tech/tags/automotive.md>), [computing](<https://devfeed.tech/tags/computing.md>), [materials](<https://devfeed.tech/tags/materials.md>), [materials-science](<https://devfeed.tech/tags/materials-science.md>), [quantum](<https://devfeed.tech/tags/quantum.md>), [science](<https://devfeed.tech/tags/science.md>)

### AI overview

IonQ describes a collaboration with an automotive manufacturer applying a hybrid quantum-classical Generative Adversarial Network to generate synthetic steel microstructure images. The approach used a Quantum-Circuit Born Machine on IonQ's Aria processor to inform a Wasserstein GAN, and the article reports higher image-quality scores than classical generative models in up to 70% of cases. The proposed use is to augment limited steel microscopy datasets for materials research and manufacturing analysis.

### Source excerpt

Working to build the world's best quantum computers to solve the world's most complex problems

## What developers need to know about generative AI

DevFeed: [What developers need to know about generative AI](<https://devfeed.tech/articles/what-developers-need-to-know-about-generative-ai-68035.md>)

Original publisher: [Read original article](<https://github.blog/ai-and-ml/generative-ai/what-developers-need-to-know-about-generative-ai/>)

Author: Damian Brady

Published: 2023-04-07T15:00:32Z

Content type: article

Language: en

Sources: [GitHub Blog](<https://devfeed.tech/sources/github-engineering.md>)

Topics: [Generative AI](<https://devfeed.tech/topics/generative-ai.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Generative Adversarial Network](<https://devfeed.tech/topics/generative-adversarial-network.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-coding-tools](<https://devfeed.tech/tags/ai-coding-tools.md>), [ai-insights](<https://devfeed.tech/tags/ai-insights.md>), [ai-ml](<https://devfeed.tech/tags/ai-ml.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [developers](<https://devfeed.tech/tags/developers.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [github-copilot](<https://devfeed.tech/tags/github-copilot.md>)

### AI overview

Generative AI systems create new text, images, audio, and code from prompts, using models such as large language models, GANs, transformers, and VAEs. The article surveys uses across industries, with emphasis on developer workflows: tools can generate or complete code, translate between programming languages, automate repetitive tasks, and support accessibility. It recommends reviewing generated code before merging and describes GitHub Copilot as a productivity aid that complements developers' work.

### Source excerpt

Generative AI has been dominating the news lately--but what exactly is it? Here's what you need to know, and what it means for developers.