探索全球前沿学术脉络

01.

arXiv (CS.AI) 2026-06-12 DOI: arXiv:2606.12797

The Containment Gap: How Deployed Agentic AI Frameworks Fail Public-Facing Safety Requirements

作者:

Md Jafrin Hossain↗Mohammad Arif Hossain↗Weiqi Liu↗Nirwan Ansari↗

arXiv:2606.12797v1 Announce Type: new Abstract: Agentic large language model systems that autonomously invoke tools, maintain persistent memory, and execute multi-step plans are increasingly deployed in public-facing domains, including government services, healthcare triage, and financial advising. We ask whether the frameworks used to build these systems provide architectural-level structural safety guarantees. Applying six containment principles derived from a compositional model of agentic architectures, we audit three dominant frameworks (LangChain, AutoGPT, and OpenAI Agents SDK) and find no native compliance in any of them. Memory integrity, a defense against one of the most prevalent vulnerability classes, is not observed in any of the three evaluated frameworks. We validate these findings empirically: in a simulated government benefits agent built on LangChain, a single memory-poisoning write induces persistent targeted corruption across all tested seeds and backends, increasing the wrongful denial rate for targeted applicants to 88.9%. Under a complex five-factor policy, the same attack preserves aggregate accuracy while increasing targeted wrongful denials by 3.5x, rendering the corruption difficult to detect through standard monitoring. We then introduce two lightweight containment mechanisms: a memory integrity validator and a policy gate, which eliminate both attack vectors with sub-millisecond overhead (

阅读与讨论 → 访问原文 →

02.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.16682

Multimodal Evaluator Preference Collapse: Cross-Modal Contagion in Self-Evolving Agents

作者:

Zewen Liu↗

When AI agents use language models to evaluate their own outputs in a feedback loop, systematic biases emerge. We show that Evaluator Preference Collapse (EPC) is dramatically amplified in multimodal settings. Using GPT-4o to evaluate DeepSeek-chat across text and visual tasks, we find that a single strategy (step_by_step) absorbs 48.4% of all weight – 3.2x the collapse observed in text-only self-evaluation – while three visual-domain strategies receive only 9.1% combined weight. We then demonstrate a novel phenomenon we term cross-modal contagion: evaluator preferences acquired on one modality transfer to and corrupt strategy selection on another. Through a four-phase isolation training paradigm, we measure contagion coefficients and document strategy inversion – the optimal strategy for a modality reverses after cross-modal exposure. A Phase 3 statistical validation across four evaluator configurations (N=53 total independent repetitions, 15,592 API calls) reveals a clear hierarchy: cross-model evaluation (GPT-4o, N=8) produces strong but symmetric bidirectional contagion (mean gamma_{T->V}=1.176, gamma_{V->T}=1.089, Delta=-0.088, p=0.575, Cohen's d=0.29); high round counts (DashScope, 50 rounds) cause collapse to single-strategy dominance (70% zero contagion); and self-evaluation provides near-complete immunity – 97% of runs (N=30, DeepSeek-chat) yield exactly zero contagion (mean gamma=0.033, 95% CI [-0.031, 0.010], p=0.642, d=0.07). No evaluator condition shows statistically significant directional asymmetry. We introduce the contagion matrix indexed by evaluator identity, release the MM-EPC experimental framework, and identify cross-model evaluator architecture as the primary risk factor for preference contagion.

阅读与讨论 → 访问原文 →

03.

arXiv (CS.AI) 2026-06-15 DOI: arXiv:2606.14200

When Should Agent Trust Be Conditional? Characterizing and Attacking Skill-Conditional Reputation in Agent Swarms

作者:

Yihan Xia↗Taotao Wang↗

arXiv:2606.14200v1 Announce Type: new Abstract: Open platforms increasingly route tasks among heterogeneous LLM agents–differing in base model, scaffold, and tool stack–whose competence varies sharply by skill: an agent excellent at one skill may be useless at another. The standard reputation approach summarizes each agent by a single global trust score, but that scalar is the wrong object here, because routing every task to the globally most-trusted agent leaves the value of specialization unclaimed. We study skill-conditional trust R(i | k)–the trust to place in agent i for a task requiring skill k, rather than one score per agent–and pose three falsifiable questions: when is conditioning worth it, how much cross-skill evidence should be borrowed, and whether that borrowing is safe. A controlled phase-diagram analysis answers the first two: conditional trust wins only in a specific regime–high agent heterogeneity, sparse per-skill evidence, and correlated skills–and the coupling strength beta that buys this data efficiency is dual-use, because the same cross-skill borrowing is also a laundering channel. On a public benchmark of 14 genuinely heterogeneous AppWorld agents, real pools land inside the beneficial regime–a small but genuine gain, with the per-skill best agent genuinely changing across skills. We then show that an attacker with cheap evidence in one skill and none in a target skill hijacks the conditional router, driving routing regret from 0 to 0.94 on a pool our zero-cost Conditional Information Value Test (CIVT) rates GREEN–while the ungated trust verdict it contaminates reads -0.06 instead of the honest +0.19. A zero-evidence gate bounds the attack but does not eliminate it; we characterize the residual cost under an explicit budget. We do not claim Sybil-resistance–we quantify the trade-off.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2606.12203

Adaptive Multi-Resolution Procedural Knowledge Compression for Large Language Models

作者:

Changyue Wang↗Weihang Su↗Qingyao Ai↗Yichen Tang↗Runzhong Qiao↗Xuancheng Li↗Min Zhang↗Yiqun Liu↗

Large language models (LLMs) are widely used to tackle complex tasks with autonomous workflows. Recently, reusable natural language skills have emerged as a popular paradigm to inject procedural knowledge into LLM applications. Since popular skills are often invoked repeatedly, placing their full text in every context significantly increases prefill cost and latency. While text compression techniques have the potential to solve this problem, most existing methods are designed to compress factual knowledge in documents instead of procedural knowledge, making them insufficient for skill compression. In this paper, we argue that an effective skill compression method should: 1) preserve logical dependencies among workflows and tool protocols, 2) enable lightweight, offline compression for frequently updated community skills, and 3) be adaptable to varying complexities across skills. To address this, we present SKIM (SKIll coMpression), an adaptive multi-resolution soft token compression framework for procedural skills. Depending on the complexity of each skill, SKIM creates different numbers of soft tokens that not only improve the efficiency of LLM inference, but also preserve the effectiveness of skill usage. Experiments indicate that SKIM compresses skills to 30 to 60 percent of their original token length while preserving task performance better than existing compression methods.We have released our code at https://github.com/bebr2/SKIM .

阅读与讨论 → 访问原文 →

05.

arXiv (CS.LG) 2026-06-17 DOI: arXiv:2606.18022

Recursive Scaling in Masked Diffusion Models

作者:

Alba Carballo-Castro↗Julianna Piskorz↗Paulius Rauba↗Mihaela van der Schaar↗Pascal Frossard↗

arXiv:2606.18022v1 Announce Type: new Abstract: Masked diffusion models (MDMs) have recently emerged as a promising paradigm for sequence generation. Scaling MDMs is conventionally achieved by increasing the parameter count or the number of denoising steps. We introduce Recursive Masked Diffusion Models (R-MDMs), which add recursive depth as a third scaling axis by repeatedly applying the same denoising transformer within each diffusion step. Recursion enables iterative refinement of the output through parameter reuse, increasing effective model depth without increasing parameter count. Across structured generation tasks, including Sudoku and Countdown, we show that R-MDMs achieve substantially improved parameter efficiency: a model with $L$ recursive iterations often matches the performance of non-recursive baselines with roughly $L\times$ more parameters. Moreover, recursive refinement can partially substitute for additional denoising steps, allowing recursive models to reach the same generation quality with fewer forward passes at inference time. These results suggest that recursive depth is a practically useful scaling mechanism for MDMs, improving both parameter efficiency and the allocation of test-time compute.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.AI) 2026-06-15 DOI: arXiv:2606.08881

Benchmarking Vision-Language-Action Models on SO-101: Failure and Recovery Analysis

作者:

Yi Yu↗Xinchuan Qiu↗

arXiv:2606.08881v2 Announce Type: replace-cross Abstract: Vision-Language-Action (VLA) models have demonstrated strong generalization in robotic manipulation, yet existing evaluations are primarily conducted in simulation or on expensive robotic platforms, leaving their robustness on affordable real-world robots largely unexplored. We present a standardized real-world benchmark for evaluating representative VLA and imitation learning policies on the low-cost SO-101 robotic platform. The benchmark comprises four representative manipulation tasks together with unified evaluation protocols, enabling systematic comparison under embodiment uncertainty. Using real-world teleoperated demonstrations, we fine-tune and evaluate $\pi_{0.5}$, SmolVLA, Wall-X, and ACT directly on the physical platform. Beyond conventional task success rates, the benchmark incorporates a structured failure taxonomy, semantic- and execution-level failure decomposition, and recovery-aware evaluation metrics to characterize policy robustness. Experimental results show that stronger pretrained VLA policies generally outperform the imitation learning baseline, although performance remains highly task-dependent under low-cost robotic deployment conditions. Execution instability emerges as the dominant failure source, while recovery capability varies substantially across architectures. These results highlight the importance of failure and recovery analysis beyond binary task success and establish SO-101 as a practical benchmark for evaluating embodied AI systems under realistic low-cost robotic deployment conditions.

阅读与讨论 → 访问原文 →

07.

arXiv (quant-ph) 2026-06-12 DOI: arXiv:2606.12766

Asymmetric quantum steering harvested near a Lorentz-violating BTZ black hole

作者:

Si-Yu Liu↗Xin-Ze Song↗Xiang-Yue Yu↗Wentao Liu↗Xiao-Li Huang↗Shu-Min Wu↗

arXiv:2606.12766v1 Announce Type: cross Abstract: We investigate the harvesting of quantum steering and its directional asymmetry between two Unruh-DeWitt detectors in a Lorentz-violating BTZ black hole spacetime. Since the detectors are located at different radial positions outside the black hole, they experience inequivalent local environments induced by gravitational redshift, causing Alice to undergo stronger effective thermal noise than Bob. Remarkably, we uncover a counterintuitive phenomenon in which the detector subjected to a higher effective temperature exhibits stronger steerability than the other one, revealing a nontrivial inversion of thermal intuition in curved spacetime. Furthermore, quantum steering survives only within a finite window of detector energy gaps and reaches its maximum within an optimal regime. We find that Lorentz violation suppresses steering most strongly near this optimal energy gap, indicating an enhanced sensitivity of maximal correlation extraction to symmetry breaking effects. Our results demonstrate that Lorentz violation acts as a geometric constraint on the quantum information capacity of spacetime, simultaneously restricting both the strength and the directionality of quantum correlations.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.CV) 2026-06-17 DOI: arXiv:2606.17362

DriveJudge: Rethinking Autonomous Driving Evaluation with Vision-Language Models

作者:

Xinglong Sun↗Kevin Xie↗Jenny Schmalfuss↗Despoina Paschalidou↗Xiuming Zhang↗Sanja Fidler↗Kashyap Chitta↗Jose M. Alvarez↗

Autonomous driving has shifted towards end-to-end policy learning, where reliable, interpretable policy evaluation is a fundamental challenge as driving quality is highly context-dependent. Commonly used rule-based driving metrics like EPDMS are interpretable but lack context-awareness, while recent VLMbased evaluations are context-aware but limited by ambiguous VLM outputs and weak physical grounding. To evaluate driving in a manner that is both interpretable and context-aware, we introduce DriveJudge. DriveJudge is a driving evaluation agent that combines rule-grounded evaluation with Vision-Language Model (VLM) reasoning and selectively invokes physically-grounded deterministic rule functions after interpreting the environmental context. To train and evaluate DriveJudge, we curate a large-scale dataset of 33,577 challenging driving samples with human annotations on whether the driving behavior is reasonable in the given scenario. With this dataset, we address the underexplored problem of driving metric evaluation, and introduce two human-aligned benchmark tasks: Driving Quality Classification and Trajectory Preference Selection. DriveJudge outperforms EPDMS for driving quality classification by 21.23 AUC, and the recent VLM-based DriveCritic for trajectory preference selection by 6.5%, setting a new standard for interpretable and precise driving evaluation.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.CV) 2026-06-19 DOI: arXiv:2606.19817

Training-Free Metrics for Synthetic Object Detection Data: A Proxy for Detector Performance

作者:

Myeongseok Nam↗Donghoon Yeo↗Seungwook Kim↗

With the recent advent of image generative models, synthetic data are increasingly being used to supplement limited real datasets for training computer vision models. However, not all synthetic datasets improve performance equally, and their effectiveness can only be assessed by training a downstream model, which is computationally expensive and time-consuming. This problem is pronounced in the task of object detection, where the required annotations are much more dense due to bounding boxes. In this paper, we propose a pre-computable metric family, dubbed Conditional-Composition Domain Match (CCDM), which serves as a proxy for the relative utility of candidate synthetic training sets for downstream detection. Experiments on the VisDrone-DET dataset show that the CCDM metric families achieve a Spearman correlation of 1.0 with the downstream performance of YOLOv8, clearly outperforming existing metrics for synthetic image evaluation.

阅读与讨论 → 访问原文 →

10.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2606.11387

Small Experiments, Cheaper Decisions: A Case Study in Staged Promotion for Micro-Pretraining

作者:

Felipe Chavarro Polania↗

Short pretraining runs can reduce experimental cost, but they can also over-promote configurations that only look strong at tiny budgets. We study an auditable staged-promotion protocol for a fixed micro-pretraining runner on two heterogeneous host blocks: Windows A100 and Linux L40S. Starting from twelve prior-screened configurations, we use staged budgets of 2 minutes, 5 minutes, 10 minutes, 60 minutes, and 12 hours, with frozen promotion rules before expensive continuations. The early screens are intentionally treated as unstable: the 5- and 10-minute rankings are host-sensitive, and the eventual 12-hour top-ranked condition is not the mean-best condition at the replicated 10-minute gate. Because seed ranges differ across stages, these changes are operational promotion evidence, not within-seed curves. A replicated 60-minute gate keeps the Staged Factorial Screening bridge reference in the promoted set, where it ranks first in all four 60-minute host-seed cells. In the final 12-hour confirmation package, the bridge condition ranks first in all four host-seed cells across two seeds; the greedy comparator does not meet the frozen 0.010 val_bpb near-equivalence rule; and the cheaper d8/ar48 (depth-8, aspect-48) sentinel does not meet the frozen 0.020 mean-gap rule. The executed 12-hour branch spends 144 GPU-hours, and the full staged protocol records 169.2 training GPU-hours including screening stages. Continuing all four 60-minute candidates would spend 192 GPU-hours, while continuing all nine replicated 10-minute candidates would spend 432 GPU-hours. The latter numbers are accounting counterfactuals for unrun continuations, not evidence that skipped candidates could not have overtaken the reference. The result is a bounded cost-allocation finding, not a claim of global optimality, capacity-normalized superiority, or superiority over adaptive hyperparameter optimization methods.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2604.13082

The Long Delay to Arithmetic Generalization: When Learned Representations Outrun Behavior

作者:

Laura Gomezjurado Gonzalez↗

arXiv:2604.13082v2 Announce Type: replace-cross Abstract: Grokking in transformers trained on algorithmic tasks is characterized by a long delay between training-set fit and abrupt generalization, but the source of that delay remains poorly understood. In encoder-decoder arithmetic models, we argue that this delay reflects limited access to already learned structure rather than failure to acquire that structure in the first place. We study one-step Collatz prediction and find that the encoder organizes parity and residue structure within the first few thousand training steps, while output accuracy remains near chance for tens of thousands more. Causal interventions support the decoder bottleneck hypothesis. Transplanting a trained encoder into a fresh model accelerates grokking by 2.75 times, while transplanting a trained decoder actively hurts. Freezing a converged encoder and retraining only the decoder eliminates the plateau entirely and yields 97.6% accuracy, compared to 86.1% for joint training. What makes the decoder's job harder or easier depends on numeral representation. Across 15 bases, those whose factorization aligns with the Collatz map's arithmetic (e.g., base 24) reach 99.8% accuracy, while binary fails completely because its representations collapse and never recover. The choice of base acts as an inductive bias that controls how much local digit structure the decoder can exploit, producing large differences in learnability from the same underlying task.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.LG) 2026-06-11 DOI: arXiv:2606.12075

Categorical Robustness Assessment for Machine Learning based Network Intrusion Detection Systems

作者:

Mayank Raj↗Nathaniel D. Bastian↗Lance Fiondella↗Gokhan Kul↗

arXiv:2606.12075v1 Announce Type: cross Abstract: Network Intrusion Detection Systems (NIDS) heavily utlize Machine Learning (ML) but ML models can be manipulated via adversarial attacks. These attacks add carefully crafted perturbations to network traffic data that leads to misclassifications. While prior work has demonstrated adversarial vulnerabilities in isolated settings, systematic cross-architecture as well as class and category of attack based comparisons under controlled attack conditions remain limited, leaving practitioners without clear guidance on which models to deploy in adversarial environments. This paper asks a simple question: what type of classifier architectures actually hold up when attackers try to manipulate the systems? We put three popular architectures through their paces: a 1D Convolutional Neural Network, a Long Short-Term Memory (LSTM) network, and a Random Forest (RF) ensemble. Using the ACI-IoT-2023 dataset (over 1.2 million samples spanning 12 attack types), we subject each model with FGSM and PGD adversarial attacks, which apply gradient-based perturbations in normalized feature space consistent with established adversarial ML evaluation protocols, at perturbation budgets ranging from $\epsilon=0.01$ to $\epsilon=0.1$. Surprisingly, Random Forest achieved near-perfect baseline accuracy (99.98\%), yet collapsed catastrophically under attack, dropping 73 percentage points at the smallest perturbation we tested. CNN, on the other hand, retained 95.5\% accuracy at $\epsilon=0.01$ and degraded gracefully as perturbations increased. LSTM fell somewhere in between. These findings flip the conventional wisdom where high baseline accuracy means nothing if a model shatters at the first sign of adversarial pressure. For practitioners deploying intrusion detection in adversarial environments, we recommend CNN-based architectures and provide scenario-specific deployment guidance.

阅读与讨论 → 访问原文 →

13.

arXiv (CS.CV) 2026-06-11 DOI: arXiv:2405.06995

Benchmarking Cross-Domain Audio-Visual Deception Detection

作者:

Xiaobao Guo↗Zitong Yu↗Nithish Muthuchamy Selvaraj↗Bingquan Shen↗Adams Wai-Kin Kong↗Alex C. Kot↗

Automated deception detection is crucial for assisting humans in accurately assessing truthfulness and identifying deceptive behavior. Conventional contact-based techniques, like polygraph devices, rely on physiological signals to determine the authenticity of an individual's statements. Nevertheless, recent developments in automated deception detection have demonstrated that multimodal features derived from both audio and video modalities may outperform human observers on publicly available datasets. Despite these positive findings, the generalizability of existing audio-visual deception detection approaches across different scenarios remains largely unexplored. To close this gap, we present the first cross-domain audio-visual deception detection benchmark, that enables us to assess how well these methods generalize for use in real-world scenarios. We used widely adopted audio and visual features and different architectures for benchmarking, comparing single-to-single and multi-to-single domain generalization performance. To further exploit the impacts using data from multiple source domains for training, we investigate three types of domain sampling strategies, including domain-simultaneous, domain-alternating, and domain-by-domain for multi-to-single domain generalization evaluation. We also propose an algorithm to enhance the generalization performance by maximizing the gradient inner products between modality encoders, named ``MM-IDGM". Furthermore, we proposed the Attention-Mixer fusion method to improve performance, and we believe that this new cross-domain benchmark will facilitate future research in audio-visual deception detection.

阅读与讨论 → 访问原文 →

14.

medRxiv (Medicine) 2026-06-10 DOI: HASH:75b616f65e8e5c14cb15637ebc2906ff

Optimisation of steatotic liver disease screening algorithm for resource-poor settings using machine learning

作者:

Mettananda↗Sivasumithran↗Ranaweera↗Madhubhashini↗Ranawaka↗Pathmeswaran↗Dassanayake↗

Background The European Association for the Study of the Liver (ESAL) - Steatotic Liver Disease (SLD) screening algorithm involves two steps; initial screening with FIB-4 followed by referral for vibration-controlled transient elastography (VCTE) in patients likely to have significant fibrosis (SF). However, VCTE is not widely available in resource-limited settings. Aim To optimise the EASL SLD screening algorithm for resource-poor settings using machine learning (ML). Methods We analysed data from 964 adults aged [≥]35 years who underwent VCTE at a tertiary referral centre in Sri Lanka between November 2024 and 2025. Multiple ML models using different methods and variable combinations were trained on 80% of the dataset and tested on the remaining 20%. Best models were selected based on performance and externally validated using data from 430 patients who underwent VCTE before November 2024. Model performance was compared with the FIB-4 using confusion matrices. Results A Random Forest model incorporating age, AST, ALT, and platelet count separately, rather than using FIB-4, outperformed. The all-variable ML model showed the best predictive performance for SF, with accuracy of 77.2%, recall of 0.762, precision of 0.778, and AUC-ROC of 0.818. The variables used in the model, in descending order of feature importance, were AST, platelet count, BMI, ALT, age, diabetes mellitus, hypertension, dyslipidaemia, sex, family history, hypothyroidism, diabetes complication and smoking. External validation demonstrated 75.1% accuracy and an AUC of 0.779. When used as the first step of the SLD screening algorithm, the all-variable ML model identified 37 (17.1%) additional true positives and reduced false-negative diagnoses by 50% compared with FIB-4. Conclusions ML-based models were more effective than the FIB-4 score as the first-line screening tool for VCTE referral, substantially improving the identification of patients with significant fibrosis in this South Asian cohort.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.CV) 2026-06-15 DOI: arXiv:2606.14556

Visual Quality Score Assessment of Large White Goods in Remanufacture with Multi-View Deformable-DETR

作者:

Paul Koch↗Vivek Chavan↗

Remanufacturing large white goods is essential for a circular economy, yet visual quality assessment remains a manual bottleneck for training and pricing. Conventional detection methods require extensive annotation and struggle with small defects in high-resolution multi-view data. We present a multi-view framework based on Deformable-DETR for automated quality scoring that aggregates information across redundant views to extract fine-grained features. To enhance robustness with limited labels, we employ self-supervised pretraining followed by supervised fine-tuning on expert-annotated scores. Additionally, a linear projection over frozen feature maps identifies regions of interest to explain model decisions. Evaluated on an industrial multi-view dataset, our approach delivers precise quality assessments while reducing reliance on manual annotation and per-part customization, enabling scalable and transparent inspection for remanufacturing lines.

阅读与讨论 → 访问原文 →

16.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2511.09465

Branching Flows: Discrete, Continuous, and Manifold Flow Matching with Splits and Deletions

作者:

Lukas Billera↗Hedwig Nora Nordlinder↗Jack Collier Ryder↗Anton Oresten↗Aron St{\aa}lmarck↗Theodor Mosetti Bj\"ork↗Ben Murrell↗

arXiv:2511.09465v4 Announce Type: replace-cross Abstract: Diffusion and flow matching approaches to generative modeling have shown promise in domains where the state space is continuous, such as image generation or protein folding & design, and discrete, exemplified by diffusion large language models. They offer a natural fit when the number of elements in a state is fixed in advance (e.g. images), but require ad hoc solutions when, for example, the length of a response from a large language model, or the number of amino acids in a protein chain is not known a priori. Here we propose Branching Flows, a generative modeling framework that, like diffusion and flow matching approaches, transports a simple distribution to the data distribution. But in Branching Flows, the elements in the state evolve over a forest of binary trees, branching and dying stochastically with rates that are learned by the model. This allows the model to control, during generation, the number of elements in the sequence. We also show that Branching Flows can compose with any flow matching base process on discrete sets, continuous Euclidean spaces, smooth manifolds, and `multimodal' product spaces that mix these components. We demonstrate this in three domains: small molecule generation (multimodal), antibody sequence generation (discrete), and protein backbone generation (multimodal), and show that Branching Flows is a capable distribution learner with a stable learning objective, and that it enables new capabilities.

阅读与讨论 → 访问原文 →

17.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2606.11695

Noise-Aware Framework for Correcting Corrupted Labels

作者:

Ha-Linh Nguyen↗Hong-Anh Nguyen↗Minh-Duc La↗Phong Lam↗Thu-Trang Nguyen↗Son Nguyen↗Hieu Dinh Vo↗

arXiv:2606.11695v1 Announce Type: cross Abstract: High-quality labeled data is essential for training reliable ML/DL models. However, real-world datasets often contain a considerable proportion of corrupted labels, which can severely degrade model performance. To address this problem, we propose CANOLA, a novel framework for correcting corrupted labels through noise-aware learning and iterative label refinement. CANOLA explicitly estimates the underlying noise distribution of the dataset and incorporates this information into the training of a noise-aware Deep Neural Network. By incorporating noise characteristics during learning, CANOLA enables the model to down-weight unreliable supervision signals and focus on trustworthy patterns, thereby improving robustness and generalization. Label correction is performed via cautious, iterative soft label refinement, in which model predictions are blended with observed labels to prevent premature or erroneous updates. This progressive refinement allows the dataset to be repaired in a stable and controlled manner. We evaluate CANOLA on six widely used datasets under realistic noisy labeling scenarios. Experimental results show that CANOLA consistently outperforms SOTA label correction methods, achieving relative improvements ranging from 19% to 52% in error reduction. Moreover, models trained on datasets corrected by CANOLA obtain substantial downstream performance gains. Even simple classifiers trained on CANOLA's corrected data can outperform complex model-centric approaches by margins of up to 67%.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.13044

No Hidden Prompts Needed! You Can Game AI Peer Review with Presentation-Only Revisions

作者:

Xu Yang↗Zhizhou Sha↗Junbo Li↗Jian Yu↗Yifan Sun↗Matthew Zhao↗Jinrui Fang↗Xinyue Guo↗Yining Wu↗Xu Hu↗Yifu Luo↗Qiang Liu↗…

As AI-generated reviews move from experimental tools into peer-review infrastructure, most robustness concerns have focused on explicit attacks such as hidden instructions and prompt injection. We study a harder and more policy-relevant failure mode: no hidden text, no prompt injection, and no changes to methods, experiments, figures, equations, proofs, or numerical results. The attacker modifies only presentation-level content, such as the abstract, contribution framing, related work, discussion, and narrative structure. We introduce adversarial repackaging: a closed-loop attack that uses AI-reviewer feedback to search for presentation-level revisions while keeping the scientific evidence fixed. Across three mainstream AI reviewers, adversarial repackaging achieves a 75.1% attack success rate and a mean score gain of +1.21/10. The effect is not explained by ordinary prose polishing. We also reveal that strategies that change how the reviewer interprets the paper, such as related-work repositioning and analytical discussion expansion, substantially outperform surface edits such as local polishing, table formatting, and algorithm boxes. Our analysis reveals two deeper structural failure modes. First, AI reviewers are easier to impress than to convince: highlighting strengths reliably increases perceived merit, while attempts to dissolve weaknesses frequently backfire. Second, AI reviewers can confuse the appearance of addressing a limitation with actually resolving it, allowing unchanged evidence to be reinterpreted as stronger scientific contribution. These results show that the deployment risk is not only malicious hidden instructions, but the emergence of paper presentation itself as an optimization surface. We release a contamination-free rolling benchmark and attack framework for testing whether AI reviewers remain anchored to scientific content under presentation-only edits.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2605.28591

Models That Know How Evaluations Are Designed Score Safer

作者:

Katharina Deckenbach↗Haritz Puerto↗Jonas Geiping↗Sahar Abdelnabi↗

The validity of AI safety evaluations depends on models behaving consistently across controlled and deployment settings. Prior work has identified test-time contextual cues, such as hypothetical scenarios, as a source of verbalized evaluation awareness and subsequent behavioral shift. In this paper, we investigate a potential explanation of this phenomenon: evaluation meta-knowledge, defined as parametric knowledge about the structural traits that characterize evaluations. Similar to dataset contamination, where benchmark exposure leads to higher performance through memorization, we hypothesize that models trained on texts describing evaluation practices may implicitly learn to recognize and respond to evaluation-like contexts, for instance, through exposure to scientific articles or social media posts about AI benchmarking. To test this, we fine-tune models on synthetic documents describing evaluation traits such as verifiable structures or moral dilemmas. Evaluating this fine-tuned model on six safety benchmarks, we find that it is significantly safer than the base model and control model. This behavioral shift persists even when restricting the analysis to responses lacking explicit verbalization of evaluation awareness. Our results demonstrate that evaluation meta-knowledge may inflate safety benchmark performance, introducing a novel confounder that is independent of explicit memorization or verbalized evaluation awareness, thus, challenging to detect. These findings have important implications for the design and interpretation of AI safety evaluations. Our code and models are available at https://github.com/compass-group-tue/arxiv2026_evaluation_meta_knowledge.

阅读与讨论 → 访问原文 →

20.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18518

PSyGenTAB: A Privacy-Preserving Framework for Synthetic Clinical Tabular Data Generation via Constrained Optimization

作者:

Arshia Ilaty↗Hossein Shirazi↗Manasi Chitale↗Kedar Hegde↗Dhanalakshmi Ramesh↗Rashmi S. Manjunath↗Amir Rahmani↗Hajar Homayouni↗

arXiv:2606.18518v1 Announce Type: cross Abstract: The development of medical AI is constrained by limited access to high-quality clinical data due to institutional silos and strict privacy regulations such as HIPAA and GDPR. Synthetic data generation offers a potential solution, but existing methods lack principled mechanisms to explicitly manage the privacy-utility trade-off, often degrading clinically meaningful patterns or risking patient re-identification. We present PSyGenTAB, a privacy-preserving generative framework that formulates synthetic healthcare data generation as a constrained optimization problem solved using the Augmented Lagrangian Method. By embedding configurable privacy constraints directly into model training, PSyGenTAB enforces minimum privacy thresholds while maximizing clinical data utility. Across multiple clinically motivated benchmarks, PSyGenTAB preserves inter-feature clinical relationships and minority-class diagnostic patterns essential for reliable health AI. Downstream evaluation using Train-on-Synthetic, Test-on-Real and Train-on-Real, Test-on-Synthetic protocols shows that models trained on synthetic data achieve performance comparable to those trained on real patient records. Privacy auditing further demonstrates reduced exact record reproduction and strong resilience to membership inference attacks. These results establish PSyGenTAB as a principled framework for balancing privacy protection and clinical utility in synthetic healthcare data, supporting secure cross-institutional AI development.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.LG) 2026-06-19 DOI: arXiv:2606.20172

Predicting gestational age at birth in the context of preterm birth from multi-modal fetal MRI

作者:

Diego Fajardo-Rojas↗Megan Hall↗Daniel Cromb↗Mary A. Rutherford↗Lisa Story↗Emma C. Robinson↗Jana Hutter↗

arXiv:2606.20172v1 Announce Type: new Abstract: Preterm birth is associated with significant mortality and a risk for lifelong morbidity. The complex multifactorial aetiology hampers accurate prediction and thus optimal care. A pipeline consisting of bespoke machine learning methods for data imputation, feature selection, and regression models to predict gestational age (GA) at birth was developed and evaluated from comprehensive multi-modal morphological and functional fetal MRI data from 333 control cases and 93 preterm birth cases. The GA at birth predictions were classified into term and preterm categories and their accuracy, sensitivity, and specificity were reported. An ablation study was performed to further validate the design of the pipeline. Performance was evaluated using stratified 10-fold cross-validation. The pipeline achieves an R2 score of 0.13 and a mean absolute error of 2.74 weeks. It also achieves a 0.77 accuracy, 0.59 sensitivity, and 0.82 specificity across folds. The predominant features selected by the pipeline include cervical length and statistics derived from placental T2* values. The confluence of fast, motion-robust and multi-modal fetal MRI techniques and machine learning prediction allowed the prediction of the gestation at birth. This information is essential for any pregnancy. To the best of our knowledge, preterm birth had only been addressed as a classification problem in the literature. Therefore, this work provides a proof of concept. Future work will increase the cohort size to allow for finer stratification within the preterm birth cohort. Our code is available at https://github.com/dfajardorojas/ml-for-preterm-birth-.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.CV) 2026-06-15 DOI: arXiv:2512.10966

Interpretable Alzheimer's Diagnosis via Multimodal Fusion of Regional Brain Experts

作者:

Farica Zhuang↗Shu Yang↗Dinara Aliyeva↗Zixuan Wen↗Duy Duong-Tran↗Christos Davatzikos↗Tianlong Chen↗Song Wang↗Li Shen↗

Accurate and early diagnosis of Alzheimer's disease (AD) is critical for effective intervention and requires integrating complementary information from multimodal neuroimaging data. However, conventional fusion approaches often rely on simple concatenation of features, which cannot adaptively balance the contributions of biomarkers such as amyloid PET and MRI across brain regions. In this work, we propose MREF-AD, a Multimodal Regional Expert Fusion model for AD diagnosis. It is a Mixture-of-Experts (MoE) framework that models mesoscopic brain regions within each modality as independent experts and employs a gating network to learn subject-specific fusion weights. Utilizing tabular neuroimaging and demographic information from the Alzheimer's Disease Neuroimaging Initiative (ADNI), MREF-AD achieves competitive performance over strong classic and deep baselines while providing interpretable, modality- and region-level insight into how structural and molecular imaging jointly contribute to AD diagnosis. The source code is available at https://github.com/PennShenLab/mref-ad.

阅读与讨论 → 访问原文 →

23.

medRxiv (Medicine) 2026-06-17 DOI: HASH:430d28f196785eb321c7c965892d44b7

Adverse Childhood Experiences Reorganise the Brain-Personality Network Across the Psychosis Spectrum

作者:

Sarti↗Cecere↗Dallenbach↗H.-L. H↗Huppi↗R. M↗Misra↗A. R↗Edkins↗Omlor↗Blom↗J. M. C↗…

Exposure to adverse childhood experiences is a pervasive risk factor for psychosis, exhibiting a linear relationship across the psychosis spectrum from subclinical schizotypal traits to schizophrenia spectrum disorders. While this association is often conceptualised within the vulnerability-stress framework, the systemic mechanisms through which childhood trauma reconfigures the brain-personality interactome remain poorly understood. We examined clinical, neuropsychological, and neuroimaging data from a sample of low- and high-schizotypy individuals, and patients with a diagnosis of schizophrenia spectrum disorder (N=120). Our aim was to map how trauma reconfigures interactions between neurobiology and schizotypal phenomenology. We adopted a mixed graphical model approach to jointly estimate conditional dependencies between childhood trauma, regional brain morphometry, and schizotypal traits across the psychosis spectrum. Our results show that childhood trauma reconfigures the brain-personality network, shifting it from a state driven by cognitive processes to one anchored in emotional (limbic) reactivity. This transition is marked by the increased influence of impulsive traits and a significant strengthening of connections within the salience network. These changes converge with a reduced thickness of the frontal executive regions, the brain's control centres, identified in our models. Collectively, our results suggest a structural phenomenological decoupling, where trauma conditioned affective circuits may bypass weakened top-down regulatory controls. These findings highlight the necessity of using integrative frameworks to capture how trauma fundamentally reshapes the relationship between the brain and schizotypal personality.

阅读与讨论 → 访问原文 →

24.

medRxiv (Medicine) 2026-06-16 DOI: HASH:62877d9fc8065569508d8fb495ef8d94

Re-evaluating the Cross-Sectional Prevalence of Severe Age-Related Hearing Loss Using Extreme Value Statistics

作者:

Bleeck↗

Standard demographic models of age-related hearing loss (presbycusis) predominantly utilize symmetric functions, such as log-normal distributions for age-binned thresholds and 4-parameter logistic curves for prevalence estimates. While these models capture early-to-moderate degradation effectively, they structurally struggle to characterize the heavy tails associated with severe clinical impairment. In this study, we present a statistical critique using a secondary analysis of the historical Medical Research Council (MRC) National Study of Hearing (1980-1986) dataset. By applying Generalized Extreme Value (GEV) distribution theory, we demonstrate that as severity increases, the underlying statistical geometry of hearing loss shifts. The asymmetric, heavy-tailed GEV distribution provides a parsimonious description of severe impairment, requiring fewer parameters than standard symmetric models. However, we explicitly acknowledge that utilizing static population data to infer progression introduces an ecological fallacy. Furthermore, the dataset's historical nature embeds unquantified generational cohort effects. We conclude that while extreme value statistics offer a compelling mathematical framework for modeling the variance of severe presbycusis, true longitudinal datasets are required to isolate physiological degradation from historical cohort variance.

阅读与讨论 → 访问原文 →

25.

arXiv (quant-ph) 2026-06-12 DOI: arXiv:2410.09226

Resourcefulness of non-classical continuous-variable quantum gates

作者:

Massimo Frigerio↗Antoine Debray↗Nicolas Treps↗Mattia Walschaers↗

arXiv:2410.09226v4 Announce Type: replace Abstract: In continuous-variable quantum computation, identifying key elements that enable a quantum computational advantage is a long-standing issue. Starting from the standard results on the necessity of Wigner negativity, we develop a comprehensive and versatile approach in which the techniques of $(s)$-ordered quasiprobabilities are exploited to provide rigorous statements on the simulability of photonic quantum circuits consisting of previously characterized gates and thereby identifying the contribution of each quantum gate to the potential achievement of quantum computational advantage. This is achieved by means of an analysis of the so-called transfer function, allowing us to highlight the resourcefulness of a gate set. As such this technique can be straightforwardly applied to current continuous-variables quantum circuits, while also constraining the tolerable amount of losses above which any potential quantum advantage can be ruled out. We use $(s)$-ordered quasiprobability distributions on phase-space to capture the non-classical features in the protocol, and focus our technique entirely on the ordering parameter $s$. This allows us to highlight the resourcefulness and robustness to loss of a universal set of unitary gates comprising three distinct Gaussian gates and any non-Gaussian unitary gate, providing important insight on the role of non-Gaussianity.

阅读与讨论 → 访问原文 →