探索全球前沿学术脉络

01.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2606.20142

RACL: Reasoning-Agent Control Layers for Continuous Metaheuristic Learning

作者:

Ant\'on Asla Manz\'arraga↗

arXiv:2606.20142v1 Announce Type: new Abstract: This paper introduces RACL, a Reasoning-Agent Control Layer for metaheuristics. RACL places a reasoning agent above an existing optimizer. The agent does not replace the optimizer and does not modify business constraints. Instead, it controls the optimizer's internal search behavior by observing operational memory, reasoning over past behavior, formulating bounded hypotheses, testing interventions, evaluating outcomes, applying guardrails, consolidating useful policies and explaining its decisions. The experiment uses vehicle routing as a testbed, but the contribution is not a new routing solver, a particular ALNS configuration or a specific set of routing rules. The contribution is the RACL method: a way for a reasoning agent to discover, validate, consolidate and explain algorithmic control rules for a metaheuristic. In the current experimental setting, RACL improves or ties the Operational Memory Policy in 21 of 21 feasible cases and improves or ties a non-reasoning Stagnation-Triggered Policy in 18 of 21 feasible cases, with an average RACL vs STP cost delta of -0.641%. In the Sevilla-9/10 runtime sample, RACL improves average cost by -8.337% versus Fixed and -1.605% versus STP without showing material computational overhead. During the proof-of-concept, Codex was used as an in-the-loop reasoning agent observing executions, interpreting logs and proposing live bounded interventions. The policy proxy was later used only to make quantitative evaluation reproducible.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18356

SafeClawBench: Separating Semantic, Audit-Evidence, and Sandbox Harm in Tool-Using LLM Agents

作者:

Yuchuan Tian↗Mengyu Zheng↗Haocheng Mei↗Ye Yuan↗Chao Xu↗Xinghao Chen↗Hanting Chen↗Yu Wang↗

arXiv:2606.18356v1 Announce Type: cross Abstract: Tool-using language-model agents introduce security failures that go beyond unsafe text: they can disclose protected objects, write persistent memory, send messages, modify databases, or trigger harmful code and tool effects. Existing evaluations often collapse these stages into a single attack success rate, making it difficult to tell whether a model merely agreed with an attacker or actually produced observable harm. We introduce SafeClawBench, a staged benchmark for tool-using agent security with 600 controlled adversarial tasks across six attack families: direct and indirect prompt injection, tool-return injection, memory poisoning, memory extraction, and ambiguity-driven unsafe inference. SafeClawBench reports three separate endpoints: semantic attack acceptance, audit-visible harm evidence, and sandbox-observed tool/state harm. Evaluating five agent endpoints under four prompt-level policies, we find that these endpoints capture different failure modes. Without additional prompt protection, semantic failure rates vary widely across models, from 9.0% to 44.2%. Audited harm evidence is narrower than semantic failure, and under a separate executable protocol some matched task identities produce sandbox harm despite passing the Semantic Core call: in a 12,000-row matched analysis, 291 of 347 observed sandbox harms occur in rows that pass the semantic check. Prompt policies change endpoint outcomes, but their effects depend on both model and protocol. SafeClawBench provides a reproducible framework for comparing agent models and prompt-policy conditions without conflating textual compliance, evidence-supported harm, and executable state changes. The open-source dataset is available at https://huggingface.co/datasets/sairights/safeclawbench.

阅读与讨论 → 访问原文 →

03.

arXiv (CS.LG) 2026-06-18 DOI: arXiv:2606.18538

Effects of sparsity and superposition on loss in simple autoencoders

作者:

Mriganka Basu Roy Chowdhury↗Eric McLaughlin Weiner↗

arXiv:2606.18538v1 Announce Type: new Abstract: One of the major difficulties in the mechanistic interpretability of neural networks is the occurrence of polysemanticity, which suggests that each neuron is typically responsible for multiple different tasks, impeding a clean interpretation of their function. The seminal paper of Elhage et al. (2022) argues that this occurs due to superposition, a phenomenon where the neural network represents distinct features as non-orthogonal directions in a lower-dimensional space, a strategy that allows much greater compression of the data without sacrificing fidelity due to the feature sparsity of input vectors. Elhage et al. (2022) empirically validates these hypotheses in a rather natural and simple autoencoder with sparse inputs. The contribution of the present work is to analyze the mathematical basis for the occurrence and optimality of superposition, while rigorously corroborating some of their findings. In particular, we provide upper and lower bounds for the L2 reconstruction loss, tight in the very sparse regime, for power activation functions. A short list of interesting open problems are also included at the end.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2606.14740

GridVQA-X: A Framework for Evaluating Multimodal Explainability Methods

作者:

Sujay Belsare↗Sudarshan Nikhil↗Sushant Kumar↗Ponnurangam Kumaraguru↗Chirag Agarwal↗

With the increasing development of Vision-Language Models, it becomes imperative that their predictions are readily explainable to relevant stakeholders. However, the field of explainability has not kept pace with the multimodal surge. While recent Multimodal Explainable AI (MxAI) methods generate explanations to attribute the interaction between different modalities, current evaluation protocols lack the ground truth required to distinguish between true cross-modal reasoning (e.g., spatial composition) and shallow cross-modal shortcuts (e.g., Bag-of-Words attribute matching). It remains unknown whether MxAI methods faithfully capture synergistic interactions or merely hallucinate reasoning on models acting as simple feature detectors. In this paper, we introduce GridVQA-X, the first diagnostic framework specifically designed to evaluate cross-modal explainability. Unlike natural datasets, GridVQA-X leverages a closed-world synthesis logic to generate unique, mathematically guaranteed explanations. We utilize this controlled environment to train paired ground-truth models on identical architectures: $M_{pure}$, which learns robust spatial-relational reasoning and $M_{spur}$, which is structurally forced to rely on cross-modal shortcuts. This behavioral divergence creates a rigorous testbed: a faithful explainer must report distinct reasoning pathways for each model. Our findings reveal that widely used methods fail to distinguish between models relying on genuine spatial-relational reasoning and those exploiting cross-modal shortcuts, highlighting a critical gap in capturing true cross-modal synergy and misrepresenting how multimodal models actually make decisions.

阅读与讨论 → 访问原文 →

05.

Nature Medicine 2026-06-08 DOI: HASH:343c619d8269c65c96f2e2be8eb5bc08

Apitegromab for lean mass preservation during tirzepatide-induced weight loss: a randomized, double-blind, placebo-controlled phase 2 trial

作者:

Richard E. Pratley↗

Loss of lean mass in proportion to total weight loss is observed with incretin mimetic therapies such as tirzepatide and has the potential to adversely affect health and function. Apitegromab is an investigational, fully human monoclonal antibody that selectively inhibits myostatin activation and is, thereby, capable of increasing muscle mass. In the randomized, double-blind, placebo-controlled phase 2 EMBRAZE study, adults with overweight or obesity (n = 102) were randomized 1:1 to receive tirzepatide plus apitegromab (10 mg kg−1) or tirzepatide plus placebo. At week 24, apitegromab resulted in a least square mean (80% confidence interval (CI)) of 1.9 (1.2−2.7) kg less lean mass loss than placebo (P = 0.001), despite similar total body weight loss between groups, representing a 54.9% retention of lean mass relative to placebo. In participants receiving apitegromab, trough concentrations of apitegromab and total latent myostatin, a pharmacodynamic marker, both increased over time and reached a plateau after approximately 16 weeks. Incidence of adverse events (AEs) (% (95% CI)) was generally similar across apitegromab-treated participants and placebo-treated participants, with 39 of 51 (76% (63−86%)) and 36 of 51 (71% (57−81%)) participants experiencing an AE, respectively. Serious adverse events (SAEs) were balanced and experienced by one of 51 (2% (0−10%)) participants in each arm. In summary, this proof-of-concept study demonstrated that selective targeting of myostatin by apitegromab was well tolerated and effective in preserving lean mass when combined with tirzepatide. ClinicalTrials.gov identifier: NCT06445075 . In the phase 2 EMBRAZE study, participants receiving tirzepatide and apitegromab lost less lean mass compared to participants receiving tirzepatide and placebo.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2604.13301

Honeypot Protocol

作者:

Najmul Hasan↗

arXiv:2604.13301v1 Announce Type: cross Abstract: Trusted monitoring, the standard defense in AI control, is vulnerable to adaptive attacks, collusion, and strategic attack selection. All of these exploit the fact that monitoring is passive: it observes model behavior but never probes whether the model would behave differently under different perceived conditions. We introduce the honeypot protocol, which tests for context-dependent behavior by varying only the system prompt across three conditions (evaluation, synthetic deployment, explicit no-monitoring) while holding the task, environment, and scoring identical. We evaluate Claude Opus 4.6 in BashArena across all three conditions in both honest and attack modes. The model achieved 100% main task success and triggered zero side tasks uniformly across conditions, providing a baseline for future comparisons with stronger attack policies and additional models.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.LG) 2026-06-12 DOI: arXiv:2606.12940

Self-Guidance: Enhancing Neural Codecs via Decoder Manifold Alignment

作者:

Xiang Li↗Yixuan Zhou↗Jingran Xie↗Zhiyong Wu↗Hui Wang↗

arXiv:2606.12940v1 Announce Type: cross Abstract: Neural speech codecs based on Vector-Quantized VAEs (VQ-VAEs) are core audio tokenizers for speech LLMs, yet their reconstruction fidelity is bottlenecked by quantization error. Modifying the quantizer or increasing model capacity are common fixes, but they complicate downstream language modeling. Our core idea is to align the decoder's internal feature manifolds when processing both the quantized tokens and their original continuous embeddings, using a lightweight feature-mapping loss. This requires minimal training overhead and no inference-time changes. Applied to XCodec2, self-guidance improves all reconstruction metrics, achieving state-of-the-art low-bitrate performance. Notably, it enables a 4x codebook reduction without fidelity loss, which downstream TTS experiments show significantly improves LLM-based synthesis by simplifying the token modeling space. Multiple statistical observations and visualizations corroborate the enhanced internal manifold alignment in the decoder. Extensive experiments confirm its generality across various inductive biases. Self-guidance thus establishes an efficient, broadly applicable method for high-fidelity neural audio coding.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.AI) 2026-06-12 DOI: arXiv:2606.12945

Learning What to Remember: A Cognitively Grounded Multi-Factor Value Model for Agentic Memory

作者:

Zhibao Chen↗Qian Cheng↗

arXiv:2606.12945v1 Announce Type: new Abstract: Long-running LLM agents accumulate interaction histories far larger than any context window, forcing a standing decision: what to encode deeply, what to forget, and what to retrieve under a fixed memory budget. Production systems answer with semantic similarity or recency – both mis-specified for the forgetting decision, which is made at consolidation time before the future query is known. We propose a multi-factor memory value function V(m)=\sum_i w_i f_i(m) over seven interpretable factors (emotional intensity, goal relevance, value alignment, self/user relevance, task utility, reliability, and usage history) drawn from cognitive psychology, whose weights are learned from a downstream objective by a gradient-free optimiser, and whose single scalar uniformly controls encoding depth, forget risk, and retrieval rank. We make a methodological point: on LongMemEval, scoring goal relevance against the held-out evaluation question saturates gold-evidence retention at \approx 0.98 – this measures retrieval, not forgetting. In the realistic blind regime, a learned multi-factor value retains 0.770 \pm 0.011 of gold evidence across 479 usable cases, versus 0.657 for uniform weights, 0.518 for the best single factor, and 0.368 for recency; every paired gap's 95% bootstrap CI is above zero, and a neural network over the same factors ties the linear model. The learned weights are interpretable – reliability, emotional intensity, and self/user relevance dominate, while query-time goal similarity is correctly down-weighted for the forgetting decision. A controlled synthetic task with planted confounds confirms the learner recovers a separating weighting (1.00 retention) where uniform weighting fails (0.62). The substrate is open-source; all experiments run on a single CPU with no API calls.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2606.15590

Unlocking Diffusion Hierarchies: Adaptive Timestep Selection for Zero-Shot Segmentation

作者:

Ramin Nakhli↗Mahesh Ramachandran↗Luca Ballan↗

Zero-shot segmentation has recently shown notable improvement by leveraging the rich visual priors in large-scale text-to-image diffusion models, such as Stable Diffusion. However, current diffusion-based methods often face limitations due to the trade-off between spatial resolution and contextual information, as well as their reliance on a single static timestep for feature extraction. To overcome these challenges, our work introduces two key advancements. First, our Contextual Similarity Maps fuse high-resolution attention maps with rich U-Net encoder features, providing both fine-grained and robust per-pixel representations. Second, we identify an emergent hierarchical semantic progression within the denoising process of various diffusion models: representations transition from part-level abstractions at earlier timesteps to object-level abstractions at later stages. Leveraging this insight, we introduce a mechanism to adaptively select the optimal timestep for each pixel. Extensive experiments demonstrate that our method consistently outperforms existing zero-shot segmentation baselines, validating the efficacy of combining contextual features with dynamic, hierarchical timestep selection.

阅读与讨论 → 访问原文 →

10.

arXiv (quant-ph) 2026-06-16 DOI: arXiv:2603.05627

Classical Explanations in (and of) General Probabilistic Theories

作者:

John Harding↗Alex Wilce↗

arXiv:2603.05627v2 Announce Type: replace Abstract: We introduce a notion of the ``explanation" of one (generalized) probabilistic model by another as particular kind of span in the category $\Prob$ of probabilistic models and morphisms. We show that explanations compose under a standard pullback construction (notwithstanding that $\Prob$ does not support arbitrary pullbacks). We then show that every locally-finite probabilistic model has a canonical, sharp classical explanation. The construction is functorial, so every locally-finite probabilistic theory has a canonical, sharp classical (though of course, usually non-local) representation.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18811

Rescaling MLM-Head for Neural Sparse Retrieval

作者:

Youngjoon Jang↗Seongtae Hong↗Jonah Turner↗Heuiseok Lim↗

arXiv:2606.18811v1 Announce Type: cross Abstract: Learned sparse retrieval (LSR) models such as SPLADE have traditionally used BERT-style masked language models as backbone encoders. A natural expectation is that replacing BERT with stronger pretrained encoders should improve retrieval effectiveness. However, we find that under standard SPLADE training recipes, backbones with large MLM-head L2 norms can suffer performance degradation and even training collapse under standard SPLADE training recipes. We identify this failure as a scale mismatch in the MLM head: SPLADE directly uses MLM-head outputs to construct sparse lexical representations, and query-document relevance is computed by an unnormalized dot product over these representations. As a result, an inflated MLM-head scale can amplify sparse activations, distort matching scores, and destabilize contrastive training under common training settings. To address this issue, we introduce a simple initialization-time correction that rescales the MLM-head projection by a constant factor before SPLADE training. This zero-cost adjustment improves training stability without modifying the model architecture or training objective. Across both in-domain and out-of-domain retrieval benchmarks, this simple correction substantially improves large-norm backbones such as ModernBERT and Ettin, turning unstable training runs into competitive sparse retrievers. In several settings, the corrected models further match or surpass the classic BERT-SPLADE baseline. These findings suggest that the bottleneck in adapting pretrained encoders to LSR is not encoder capacity alone, but the calibration of the MLM-head scale used to construct sparse lexical representations.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.CV) 2026-06-12 DOI: arXiv:2606.06113

Where, What, Why, and Importance: Structured Defect Grounding for Text-to-Image Feedback

作者:

Huaisong Zhang↗Hao Yu↗Yuxuan Zhang↗Jiahe Wang↗Xinrui Chen↗Haoxiang Cao↗Feng Lu↗Wendong Zhang↗Changqian Yu↗Chun Yuan↗

Despite generating increasingly photorealistic images, text-to-image (T2I) models still exhibit localized, subtle, and structurally complex failures. Diagnosing these failures requires instance-level feedback that answers where a defect occurs, what type it is, why it is defective, and its importance to overall image quality. While recent dense-feedback methods move beyond scalar supervision, their heatmap-centric representations still formulate diagnosis as pixel-field regression, making it difficult to localize variable-cardinality defects and bind semantic reasons to individual failures. To address this representation bottleneck, we propose Structured Defect Grounding (SDG), which casts T2I diagnosis as structured set prediction by modeling each defect as a (location, type, reason, importance) tuple. To make this formulation trainable and measurable, we introduce SDG-30K, a 30K-image dataset with box-grounded annotations across four modern T2I generators, together with a dedicated evaluation protocol, SDG-Eval. Building on this structured representation, we further present a diagnosis-to-alignment framework in which a Vision-Language Model (VLM) serves as the SDG detector, and BoxFlow-GRPO converts predicted defect sets into box-derived, importance-weighted spatial rewards for diffusion model alignment. Extensive experiments show that our SDG detector outperforms leading proprietary VLMs on structured defect grounding, while SDG-guided rewards consistently improve T2I alignment and support localized image refinement. These results establish SDG as a unified, instance-level interface for diagnosing, evaluating, and enhancing modern generative models.

阅读与讨论 → 访问原文 →

13.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2502.10389

Region-Adaptive Sampling for Diffusion Transformers

作者:

Ziming Liu↗Yifan Yang↗Chengruidong Zhang↗Yiqi Zhang↗Lili Qiu↗Yang You↗Yuqing Yang↗

Diffusion models (DMs) have become the leading choice for generative tasks across diverse domains. However, their reliance on multiple sequential forward passes significantly limits real-time performance. Previous acceleration methods have primarily focused on reducing the number of sampling steps or reusing intermediate results, failing to leverage variations across spatial regions within the image due to the constraints of convolutional U-Net structures. By harnessing the flexibility of Diffusion Transformers (DiTs) in handling variable number of tokens, we introduce RAS, a novel, training-free sampling strategy that dynamically assigns different sampling ratios to regions within an image based on the focus of the DiT model. Our key observation is that during each sampling step, the model concentrates on semantically meaningful regions, and these areas of focus exhibit strong continuity across consecutive steps. Leveraging this insight, RAS updates only the regions currently in focus, while other regions are updated using cached noise from the previous step. The model's focus is determined based on the output from the preceding step, capitalizing on the temporal consistency we observed. We evaluate RAS on Stable Diffusion 3 and Lumina-Next-T2I, achieving speedups up to 2.36x and 2.51x, respectively, with minimal degradation in generation quality. Additionally, a user study reveals that RAS delivers comparable qualities under human evaluation while achieving a 1.6x speedup. Our approach makes a significant step towards more efficient diffusion transformers, enhancing their potential for real-time applications.

阅读与讨论 → 访问原文 →

14.

arXiv (CS.CV) 2026-06-17 DOI: arXiv:2606.17188

Not Truly Multilingual: Script Consistency as a Missing Dimension in VLM Evaluation

作者:

Prabhjot Singh↗Bhushan Pawar↗Madhu Reddiboina↗Rajvee Sheth↗

Current multilingual evaluations for Vision-Language Models (VLMs) assume a one-to-one mapping between language and orthography, overlooking billions of users of multi-script languages. We introduce PuMVR (Punjabi Multimodal Visual Reasoning), a benchmark of 1,000 strictly parallel image-text instances across Punjabi's three active scripts: Gurmukhi, Shahmukhi, and Roman. Evaluating 10 state-of-the-art VLMs, we expose a substantial and systematic Script Gap. Models frequently solve visual tasks in one script while failing identical tasks in another, with accuracy deltas reaching 16%. Crucially, visual input boosts absolute performance uniformly yet does not close the orthographic gap. Furthermore, cross-script in-context transfer is highly brittle, exposing script-locked knowledge representation. Supported by McNemar tests across all script pairs, our findings demonstrate that current "multilingual" VLMs are not truly multi-script. We propose the Script Consistency Rate (SCR), which falls as low as 24.8% on our benchmark, as a mandatory metric for script-agnostic evaluation to ensure equitable AI access. Data and code are available at: https://github.com/prabhjotschugh/Not-Truly-Multilingual-PuMVR.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2603.04219

ZeSTA: Zero-Shot TTS Augmentation with Domain-Conditioned Training for Data-Efficient Personalized Speech Synthesis

作者:

Youngwon Choi↗Jinwoo Oh↗Hwayeon Kim↗Hyeonyu Kim↗

arXiv:2603.04219v2 Announce Type: replace-cross Abstract: We investigate the use of zero-shot text-to-speech (ZS-TTS) as a data augmentation source for low-resource personalized speech synthesis. While synthetic augmentation can provide linguistically rich and phonetically diverse speech, naively mixing large amounts of synthetic speech with limited real recordings often leads to speaker similarity degradation during fine-tuning. To address this issue, we propose ZeSTA, a simple domain-conditioned training framework that distinguishes real and synthetic speech via a lightweight domain embedding, combined with real-data oversampling to stabilize adaptation under extremely limited target data, without modifying the base architecture. Experiments on LibriTTS and an in-house dataset with two ZS-TTS sources demonstrate that our approach improves speaker similarity over naive synthetic augmentation while preserving intelligibility and perceptual quality. Audio samples are available on our web page.

阅读与讨论 → 访问原文 →

16.

arXiv (math.PR) 2026-06-11 DOI: arXiv:2411.08653

Hilbert space embeddings of independence tests and interaction measures of several variables

作者:

Jean Carlo Guella↗

arXiv:2411.08653v2 Announce Type: replace-cross Abstract: We present a unified theoretical framework for kernel-based measures of dependence on product spaces. Building on the ideas underlying distance covariance, distance multivariance, and the Hilbert-Schmidt Independence Criterion (HSIC), we define a new family of kernels on an $n$-fold Cartesian product, termed positive definite independent of order $k$ (PDI$_{k}$ kernels). These kernels extend the concepts of positive definite and conditionally negative definite kernels to higher orders and provide the foundation for generalized independence and interaction tests, such as the generalized Lancaster interaction of order $k$ ($\Lambda_{k}^{n}$), and the Streitberg interaction ($\Sigma$). Our analysis focuses on the continuous setting, where we prove a Kernel Mean Embedding Theorem for PDI$_{k}$ kernels and establish the corresponding integrability restrictions. Based on these results, we characterize how the Kronecker products of PDI kernels behave.

阅读与讨论 → 访问原文 →

17.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2606.20520

Sovereign Execution Brokers: Enforcing Certificate-Bound Authority in Agentic Control Planes

作者:

Jun He↗Deying Yu↗

arXiv:2606.20520v1 Announce Type: cross Abstract: Autonomous agents are increasingly connected to cloud, deployment, and data-control workflows, but production mutation authority should not reside inside non-deterministic reasoning processes. Existing access-control mechanisms authorize identities, while assurance layers certify proposed actions; neither alone provides a mandatory enforcement point for certified authority at the moment of mutation. This paper introduces the Sovereign Execution Broker (SEB), a runtime enforcement boundary for certificate-bound agentic infrastructure. SEB consumes certificates issued by the Sovereign Assurance Boundary (SAB), verifies that the requested mutation matches the certified execution contract, checks validity windows, policy epochs, revocation epochs, and live-state drift, mints scoped execution identity, invokes infrastructure APIs, and records signed decision and outcome records. By separating proposal, admission, and execution, SEB turns certified authority into a short-lived, revocable, auditable runtime capability, provided that production mutation APIs reject non-broker identities. We present the SEB execution model, certificate and replay-verification predicates, scoped identity semantics, bypass-prevention deployment patterns, failure behavior, and a concrete prototype implementation. We evaluate the prototype on AWS and Kubernetes clusters, measuring latency overheads, revocation propagation, drift detection, and security under fault injection.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.CL) 2026-06-18 DOI: arXiv:2606.18471

Possible or Definite? A Benchmark for Evaluating Diagnostic Uncertainty Preservation in Clinical Text

作者:

Hongbo Du↗Zixin Lu↗Jiaming Qu↗

Large language models (LLMs) are increasingly used for clinical text tasks such as summarization and revision. While most studies evaluate the fluency and coherence of LLM-generated text, whether LLMs correctly preserve diagnostic uncertainty remains underexplored. In clinical practice, phrases such as ``possible pneumonia'' communicate the strength of available evidence and directly guide decisions about follow-up testing and treatment. Altering these uncertainty expressions can change the clinical meaning entirely. In this paper, we systematically evaluated this problem in two steps. First, we constructed a benchmark of 1,200 clinical documents with 9,184 uncertainty annotations across five levels. Second, we evaluated three LLMs on this benchmark. Our results show that (1) LLMs preserve the original uncertainty cues poorly, often less than half the time; (2) LLMs struggle with nuanced distinctions between adjacent levels. This work reveals a failure mode not captured by standard evaluation metrics and provides implications for the safe deployment of LLMs in clinical workflows.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.15788

GAS-Leak-LLM: Genetic Algorithm-Based Suffix Optimization for Black-Box LLM Jailbreaking

作者:

Aman Anifer↗Vignesh Kumar Kembu↗Vishnu M↗Antonino Nocera↗Vinod P↗Amal Murali PK↗Akshay S Rajan↗

arXiv:2606.15788v1 Announce Type: cross Abstract: Large Language Models (LLMs) constitute pivotal components within the AI-dominated information technology ecosystem. To mitigate risks associated with harmful or policy-violating outputs, commercial systems employ advanced alignment strategies and multi-layered content moderation mechanisms. Despite these safeguards, recent research has demonstrated that LLMs remain vulnerable to adversarial manipulation, particularly through jailbreaking and prompt injection techniques. In this work, we propose GAS-Leak-LLM a novel jailbreaking attack based on a genetic algorithm that systematically evolves adversarial suffix to bypass safety constraints. Operating in a strict black-box setting, our method requires no access to model parameters or internals, thereby reflecting realistic threat scenarios in deployed systems. Through the iterative application of selection, mutation, and crossover heuristics, the framework systematically explores the discrete prompt space to identify high-fitness adversarial suffixes. Empirical findings reveal critical shortcomings in existing safety enforcement mechanisms and confirm the effectiveness and practical viability of the proposed attack.

阅读与讨论 → 访问原文 →

20.

arXiv (math.PR) 2026-06-18 DOI: arXiv:2510.17629

Formation of clusters and coarsening in weakly interacting diffusions

作者:

Nicolai Gerber↗Rishabh S. Gvalani↗Martin Hairer↗Grigorios A. Pavliotis↗Andr\'e Schlichting↗

arXiv:2510.17629v3 Announce Type: replace-cross Abstract: This paper studies the clustering behavior of weakly interacting diffusions under the influence of sufficiently localized attractive interaction potentials on the one-dimensional torus. We describe how this clustering behavior is closely related to the presence of discontinuous phase transitions in the mean-field PDE. For local attractive interactions, we employ a new variant of the strict Riesz rearrangement inequality to prove that all global minimizers of the free energy are either uniform or single-cluster states, in the sense that they are symmetrically decreasing. We analyze different timescales for the particle system and the mean-field (McKean-Vlasov) PDE, arguing that while the particle system can exhibit coarsening by both coalescence and diffusive mass exchange between clusters, the clusters in the mean-field PDE are unable to move and coarsening occurs via the mass exchange of clusters. By introducing a new model for this mass exchange, we argue that the PDE exhibits dynamical metastability. We conclude by presenting careful numerical experiments that demonstrate the validity of our model.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2509.22808

ArFake: A Robust Framework for Multi-Dialect Arabic Speech Spoofing Detection Benchmark

作者:

Mohamed Elsetohy↗Alhassan Ehab↗Ali Mekky↗Besher Hassan↗Shady Shehata↗

With the rise of generative text-to-speech models, distinguishing between real and synthetic speech has become challenging, especially for Arabic that have received limited research attention. Most spoof detection efforts have focused on English, leaving a significant gap for Arabic and its many dialects. In this work, we introduce the first multi-dialect Arabic spoofed speech dataset. To evaluate the difficulty of the synthesized audio from each model and determine which produces the most challenging samples, we aimed to guide the construction of our final dataset either by merging audios from multiple models or by selecting the best-performing model, we conducted an evaluation pipeline that included training classifiers using two approaches: modern embedding-based methods combined with classifier heads; classical machine learning algorithms applied to MFCC features; and the RawNet2 architecture. The pipeline further incorporated the calculation of Mean Opinion Score based on human ratings, as well as processing both original and synthesized datasets through an Automatic Speech Recognition model to measure the Word Error Rate. Our results demonstrate that FishSpeech outperforms other TTS models in Arabic voice cloning on the Casablanca corpus, producing more realistic and challenging synthetic speech samples. However, relying on a single TTS for dataset creation may limit generalizability.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.LG) 2026-06-11 DOI: arXiv:2601.21817

A Judge-Aware Ranking Framework for Evaluating Large Language Models without Ground Truth

作者:

Mingyuan Xu↗Xinzi Tan↗Jiawei Wu↗Doudou Zhou↗

arXiv:2601.21817v3 Announce Type: replace-cross Abstract: Evaluating large language models (LLMs) on open-ended tasks without ground-truth labels is increasingly done via the LLM-as-a-judge paradigm. A critical but under-modeled issue is that judge LLMs differ substantially in reliability; treating all judges equally can yield biased leaderboards and misleading uncertainty estimates. More data can make evaluation more confidently wrong under misspecified aggregation. We propose a judge-aware ranking framework that extends the Bradley-Terry-Luce model by introducing judge-specific discrimination parameters, jointly estimating latent model quality and judge reliability from pairwise comparisons without reference labels. We establish identifiability up to natural normalizations and prove consistency and asymptotic normality of the maximum likelihood estimator, enabling confidence intervals for score differences and rank comparisons. Across multiple public benchmarks and a newly collected dataset, our method improves agreement with human preferences, achieves higher data efficiency than unweighted baselines, and produces calibrated uncertainty quantification for LLM rankings.

阅读与讨论 → 访问原文 →

23.

medRxiv (Medicine) 2026-06-11 DOI: HASH:1f312bb66705de39dcd7d9b8536d1724

Development of iADJUST: a theory-informed, patient co-designed digital psychological intervention for adjustment in chronic kidney disease

作者:

Schmill↗Hudson↗Greenwood↗Chilcot↗

Background: Psychological distress is common in chronic kidney disease (CKD) and is associated with reduced quality of life, treatment non-adherence, and worse clinical outcomes. Distress in CKD is also linked to difficulties adjusting to the demands of illness management. Despite this, psychological support remains inconsistently integrated within kidney care pathways, and existing interventions often lack clear theoretical specification and explicit targeting of mechanisms underpinning adjustment to CKD. Objectives: To describe the systematic development of iADJUST, a theory-informed patient co-designed digital psychological intervention targeting key cognitive and behavioural mechanisms involved in adjustment to CKD. Methods: Intervention development was guided by the Medical Research Council framework for complex interventions. A structured, iterative process integrated empirical evidence, psychological theory, and patient and public involvement and engagement. The Common-Sense Model of Self-Regulation and cognitive behavioural theories informed the identification of modifiable maintaining mechanisms associated with adjustment to CKD. Intervention components were mapped onto these mechanisms and refined through co-design with people living with CKD. Results: iADJUST is a six-session self-guided digital psychological intervention delivered over 12 weeks and supplemented by therapist contact. The intervention targets illness-related uncertainty, fatigue-related activity dysregulation, catastrophic what-if thinking, self-critical evaluation, and behavioural withdrawal. It integrates psychoeducation, cognitive and behavioural strategies, maintenance planning, and elements from acceptance and commitment therapy and compassion-focused approaches. Content is delivered through video, audio, and guided tasks and activities. Conclusion: iADJUST provides a theory-informed, evidence-based psychological intervention for CKD explicitly mapping intervention components to maintaining cognitive and behavioural mechanisms implicated in adjustment. Feasibility evaluation is underway.

阅读与讨论 → 访问原文 →

24.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18268

Towards Multi-Agent-Simulation-Based Community Note Evaluation

作者:

Changxi Wen↗Shuning Zhang↗Bohao Chu↗Yuwei Chuai↗Hui Wang↗Dai Shi↗Xin Yi↗Hewu Li↗

arXiv:2606.18268v1 Announce Type: cross Abstract: Community-based fact-checking that relies on cross-consensus is expanding rapidly on social media platforms. However, the delay and low-ratio of cross-consensus community fact-checks rated by human contributors remains a significant challenge. To address this, we first created ComRate, a large-scale dataset comprising 2.5 million community notes and over 209 million ratings sourced from $\mathbb{X}$. We then propose MultiCom, a persona-guided multi-agent rating framework for community note evaluation. MultiCom simulates diverse rater population by clustering contributors in a matrix-factorized rater space and prompting persona agents to generate structured assessments based on the official community notes rating schema. These agents output structured and explainable judgments, such as confidence, agreement signals and reasons. An out-of-fold calibrated aggregation algorithm combines features such as raw votes and diagnostic reason signals for reliable prediction. Extensive evaluations demonstrate that MultiCom outperforms alternative methods, achieving an average accuracy of 84.7% (balanced accuracy 68.3%, macro-F1 60.1%) on the evaluation set.

阅读与讨论 → 访问原文 →

25.

arXiv (math.PR) 2026-06-17 DOI: arXiv:2510.00343

Limit theorems for descents and inversions of shelf-shuffles

作者:

Alexander Clay↗

arXiv:2510.00343v2 Announce Type: replace Abstract: We prove central limit theorems for the number of descents and inversions of permutations produced by shelf-shuffles. These are a model for casino card shuffling machines. We show the asymptotic normality of the number of descents in two limiting regimes depending on the ratio of cards to shelves. On the other hand, we study the inversions by employing a modification of the techniques from Islak's analysis of the statistics of riffle shuffles. In particular, we obtain a bound for the rate of convergence for inversions that is independent of the number of shelves.

阅读与讨论 → 访问原文 →