论文广场 - AcademicHub

01.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.17029

DEEPRUBRIC: Evidence-Tree Rubric Supervision for Efficient Reinforcement Learning of Deep Research Agents

作者:

Minghang Zhu ↗Chuyang Wei ↗Junhao Xu ↗Yilin Cheng ↗Zhumin Chen ↗Jiyan He ↗

Deep research agents synthesize long-form reports by searching and reasoning over retrieved evidence. Reinforcement learning with rubric-based rewards improves these agents by optimizing them against checkable criteria that translate report quality into reward signals, but its efficiency depends on whether those criteria reliably capture the task scope and evidence needs. Most existing studies ask an LLM to generate rubrics for a given query, but when the model fails to infer the underlying information needs, the generated rubrics may be incomplete and reduce RL efficiency. To obtain more reliable query–rubric supervision, we introduce DeepRubric, a data construction framework that reverses this process: instead of inferring evaluation criteria for a given query, it first determines what an evidence-backed report should be evaluated on and then synthesizes aligned query–rubric pairs from those evaluation targets. Starting from a sampled seed topic, DeepRubric builds an evidence tree by recursively expanding evidence-backed sub-questions, whose leaves serve as atomic and verifiable evaluation targets. It then uses the evidence tree to synthesize the training query and rubrics, ensuring that the reward evaluates exactly the information requested by the query. Using DeepRubric, we construct 9K query–rubric supervision examples and train DeepRubric-8B with rubric-based GRPO, achieving comparable performance to prior open state-of-the-art deep research models across three benchmarks with roughly 13x fewer RL GPU-hours.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2606.11238

Artificial Intelligence in Ship Finance: Applications, Opportunities, and a Case Study in AI-Augmented Loan Origination

作者:

Lasse Dierich ↗Orestis Schinas ↗

arXiv:2606.11238v1 Announce Type: cross Abstract: Ship finance is a data-intensive and document-heavy segment of asset-based lending, requiring the integration of financial, technical, contractual, and regulatory information from heterogeneous and largely unstructured sources. Increasing environmental regulation and ESG reporting requirements are adding further complexity to underwriting and loan-origination processes. Recent advances in artificial intelligence (AI), particularly large language models (LLMs), create new opportunities for processing and analysing such information. This paper reviews potential applications of AI in ship finance, with a particular focus on LLM-based systems for document comprehension, information extraction, and workflow automation. We present ShipFinance.ai, a modular agentic architecture to support loan application workflows in ship finance. The proposed system combines an LLM-based extraction module, financial analysis components, external maritime data services, and a controlled document-generation module with a chatbot interface to support the preparation of standardized financing applications. The paper discusses the key challenges for using such models in production. We argue that AI-assisted systems can support maritime finance professionals in managing increasingly complex information and reporting requirements.

阅读与讨论 → 访问原文 →

03.

arXiv (CS.CV) 2026-06-19 DOI: arXiv:2512.00850

Smol-GS: Compact Representations for Abstract 3D Gaussian Splatting

作者:

Haishan Wang ↗Mohammad Hassan Vali ↗Arno Solin ↗

We present Smol-GS, a novel method for learning compact representations for 3D Gaussian Splatting (3DGS). Our approach learns highly efficient splat-wise features to model 3D space, which capture abstracted cues, including color, opacity, transformation, and material properties. We propose octree-derived positional encoding, which explicitly models spatial locality and enhances representation efficiency. We further apply entropy-based compression to exploit feature redundancy and compress splat coordinates using a recursive voxel hierarchy. This design enables orders-of-magnitude reduction in storage while preserving representation flexibility. Smol-GS achieves state-of-the-art compression performance on standard benchmarks with high-level rendering quality.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.LG) 2026-06-12 DOI: arXiv:2606.10069

Using Seismic Statistical Features and VQ-VAE to Improve Spatiotemporal Seismicity Predictability

作者:

Wei Quan ↗Denise Gorse ↗

arXiv:2606.10069v2 Announce Type: replace Abstract: In this paper we build upon a previous study in which we demonstrated, using XGBoost and earthquake catalogue data from Japan and Chile, that a set of 60 seismic statistical features (SSFs) had much greater predictive value than a set of 428 generic time series features from the tsfresh package. We here extend this previous work in two key ways, focusing on data from Japan as a large dataset is necessary in order to allow for the training of a deep learning (autoencoder) model. First, we move from whole-region prediction (considering, for each candidate event, the likelihood of an event M $\geq$ 5.0 anywhere in the region in the next 15 days) to localised predictions in which both the region of feature computation and the region of prediction are restricted to a circle of radius 24 km around the candidate event, and we show that performance remains excellent, similar to our previous whole-region study for the same area. Second, we here couple this proven set of SSFs, based on one-dimensional (catalogue) data, with a novel feature based on two-dimensional seismic maps, obtained by training a VQ-VAE model to reproduce such maps as output and identifying a measure of its error in doing so with a localised build-up of crustal stress. We show that while localised prediction based on SSFs can be effective alone, with test AUC values as high as those obtained in the case of Japan in our previous whole-region study, the inclusion of the new natively-spatial VQ-VAE-derived feature, top-ranked by SHAP analysis, can enhance performance and additionally appears to near-wholly replace the traditionally-computed $b$-value in terms of feature usage.

阅读与讨论 → 访问原文 →

05.

arXiv (CS.CL) 2026-06-17 DOI: arXiv:2606.18203

RubricsTree: Scalable and Evolving Open-Ended Evaluation of Personal Health Agents across Health Memory and Medical Skills

作者:

Weizhi Zhang ↗Zechen Li ↗Hamid Palangi ↗Ben Graef ↗A. Ali Heydari ↗Simon A. Lee ↗Salman Rahman ↗Ray Luo ↗Zeinab Esmaeilpour ↗Erik Schenck ↗Chloe Zhang ↗Yamin Li ↗…

The LLM-empowered personal health agents with user health (sensor) metrics have offered a promising pathway to alleviate global disparities in healthcare access. However, large-scale clinical deployment remains constrained by an open-ended evaluation bottleneck: physician annotation is reliable but costly and unscalable, while LLM-as-a-judge evaluators are scalable but subjective, inconsistent, and sometimes clinically misaligned. We introduce RubricsTree, a scalable evaluation framework with an expert-aligned hierarchical taxonomy of over 100 atomic, clinically-verifiable Boolean rubrics, evolving from the insights of 4,000 real user queries through an iterative human-in-the-loop curation protocol with an expertise panel led by an experienced physician. A context-aware adaptive router activates only the relevant auto-weighted rubric subset per query, providing the throughput needed for scalable evaluation with expert-aligned quality. Through a systematic meta-evaluation, we show that RubricsTree (i) substantially exceeds a strong large-scale evaluation baseline in expert alignment on challenging open-ended queries; (ii) reliably penalizes contextually degraded responses; and (iii) when used as structured instructions, text feedback, or training rewards for performance optimization, yields up to ~66% relative gains on HealthBench for Gemini, GPT, and Qwen model families. RubricsTree thus provides a scalable, auditable, and evolving evaluation infrastructure required for the continuous optimization of product-level personal healthcare AI.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2606.14958

MVEB: Massive Video Embedding Benchmark

作者:

We introduce the Massive Video Embedding Benchmark (MVEB), a 23-task benchmark for video embeddings spanning classification, zero-shot classification, clustering, pair classification, retrieval, and video-centric question answering. We evaluate 33 models and find that no single model dominates: MLLM-based embeddings lead on classification, clustering, pair classification, and QA; multimodal binding leads on retrieval and zero-shot classification; generative MLLMs without contrastive adaptation collapse on cross-modal tasks. Paired video-only vs. audio+video evaluations show that audio's contribution depends on dataset annotation provenance: audio helps when labels were produced from both modalities and hurts when they were produced from visuals alone, a six-point gap consistent across model families. MVEB is derived from MVEB+, a 184-task pool, and is designed to maintain task diversity while reducing evaluation cost. It integrates into the MTEB ecosystem for unified evaluation across text, image, audio, and video. We release MVEB and all 184 tasks along with code and a leaderboard at https://github.com/embeddings-benchmark/mteb.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.AI) 2026-06-17 DOI: arXiv:2606.18075

A Unified Framework for Context-Aware and Relation-Aware Graph Retrieval-Augmented Generation

作者:

Haoyang Zhong ↗Yifei Sun ↗Antong Zhang ↗Chunping Wang ↗Lei Chen ↗Yang Yang ↗

arXiv:2606.18075v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) has emerged as a paradigm for enhancing large language models (LLMs) with external knowledge, yet existing graph-based methods face a fundamental limitation: entity-centric and chunk-centric approaches operate on representations anchored to original text without true knowledge fusion. While entity-centric methods connect logically related content and chunk-centric methods preserve context, both retrieve information separately through similarity search, missing emergent understanding from their synthesis. In this paper, we propose HyGRAG, a hierarchical graph RAG framework that transcends source documents by addressing three core challenges: constructing summaries that genuinely integrate contextual and relational information, leveraging these synthesized representations to access emergent knowledge during retrieval, and efficiently updating hierarchical structures for dynamic corpora. Specifically, we design hierarchical index structures over hybrid graphs with both chunk and entity nodes, then iteratively cluster them and generate LLM-based summaries. Then, we design context and relation-aware retrieval that searches across all abstraction levels while expanding through community membership. Moreover, we enable dynamic knowledge update through attachment-based algorithms with only local re-summarization. Experimental results show that HyGRAG improves the average accuracy of multi-hop reasoning tasks by 9.7%, while maintaining reasonable efficiency.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2605.18421

EvoMemBench: Benchmarking Agent Memory from a Self-Evolving Perspective

作者:

Yuyao Wang ↗Zhongjian Zhang ↗Mo Chi ↗Kaichi Yu ↗Yuhan Li ↗Miao Peng ↗Bing Tong ↗Chen Zhang ↗Yan Zhou ↗Jia Li ↗

Recent benchmarks for Large Language Model (LLM) agents mainly evaluate reasoning, planning, and execution. However, memory is also essential for agents, as it enables them to store, update, and retrieve information over time. This ability remains under-evaluated, largely because existing benchmarks do not provide a systematic way to assess memory mechanisms. In this paper, we study agent memory from a self-evolving perspective and introduce EvoMemBench, a unified benchmark organized along two axes: memory scope (in-episode vs. cross-episode) and memory content (knowledge-oriented vs. execution-oriented). We compare 15 representative memory methods with strong long-context baselines under a standardized protocol. Results show that current memory systems are still far from a general solution: long-context baselines remain highly competitive, memory helps most when the current context is insufficient or tasks are difficult, and no single memory form works consistently across all settings. Retrieval-based methods remain strong for knowledge-intensive settings, whereas procedural and long-term memory methods are more effective for execution-oriented tasks when their stored experience matches the task structure. We hope EvoMemBench facilitates future research on more effective memory systems for LLM-based agents. Our code is available at https://github.com/DSAIL-Memory/EvoMemBench.

阅读与讨论 → 访问原文 →

09.

PLOS Medicine 2026-05-13 DOI: HASH:09e1b7c2032e6adf96b8b52c7c1c674e

Contribution of nosocomial transmission to <i>Klebsiella pneumoniae</i> neonatal sepsis in Africa and South Asia: An observational study of infection clusters inferred from pathogen genomics and temporal data

作者:

Erkison Ewomazino Odih ↗

by Erkison Ewomazino Odih, Jabir A. Abdulahi, Anne V. Amulele, Matthew Bates, Eva Heinz, Weiming Hu, Kajal Jain, Rindidzani Magobo, Courtney P. Olwagen, John M. Tembo, Tolbert Sonda, Jonathan Strysko, Caroline C. Tigoi, Kyle Bittinger, Jennifer Cornick, Ebenezer Foster-Nyarko, Wilson Gumbi, Steven M. Jones, Chileshe L. Musyani, Carolyn M. McGann, Ahmed M. Moustafa, Patrick Musicha, James C. L. Mwansa, Moreka L. Ndumba, Thomas D. Stanton, Donwilliams O. Omuoyo, Oliver Pearse, Laura T. Phillips, Paul J. Planet, Charlene M. C. Rodrigues, Fatou Secka, Kirsty Sands, Erin Theiller, Allan M. Zuza, Sulagna Basu, Grace J. Chan, Kenneth C. Iregbu, Jean-Baptiste Mazarati, Semaria Solomon Alemayehu, Timothy R. Walsh, Rabaab Zahra, Angela Dramowski, Sombo Fwoloshi, Appiah-Korang Labi, Lola Madrid, Noah Obeng-Nkrumah, David Ojok, Boaz D. Wadugu, Andrew C. Whitelaw, Anudita Bhargava, Atul Jindal, Ramesh K. Agarwal, Alexander M. Aiken, James A. Berkley, Susan E. Coffin, Nicholas A. Feasey, Nelesh P. Govender, Davidson H. Hamer, Shabir A. Madhi, Mari Jeeva Sankar, Kelly L. Wyres, Kathryn E. Holt Background Klebsiella pneumoniae is the leading cause of sepsis among neonates in low- and middle-income countries (LMICs) in Africa and Asia, contributing substantially to the overall burden of antimicrobial-resistant infections and mortality among neonates globally. Pathogen sequencing has been used to investigate case clusters and confirm nosocomial transmission in a small number of neonatal units. Here we utilise pathogen sequence data to estimate the fraction of K. pneumoniae neonatal sepsis attributable to nosocomial transmission in African and South Asian countries. Methods and findings We estimated the proportion of invasive K. pneumoniae disease involved in nosocomial transmission clusters in a given neonatal unit, using single-linkage clustering based on pairwise temporal and genetic distances estimated from bacterial whole-genome sequences aggregated from 10 contributing studies. Analysing 1,523 K. pneumoniae isolates from 27 units in 13 countries in Africa and South Asia between 2013 and 2023, we inferred 156 nosocomial transmission clusters, ranging from 2 to 188 neonates each (83 of the clusters comprised ≥3 cases). Overall, we estimated that 1,035 neonatal infections (68.0%) were part of nosocomial transmission clusters. Excluding the first infection in each cluster as a potential index case, we estimate at least 879 (57.7%) infections were acquired via nosocomial transmission. Sensitivity analyses showed that results were robust to the choice of genetic distance estimation methods and thresholds used to define clusters, and cluster estimates were stable over temporal distance thresholds ranging from 2 to 8 weeks. Isolates were mostly extended-spectrum beta-lactamase (ESBL) producers (90.9%) and included 172 multi-locus sequence types (STs). Fourteen STs, including several globally recognised multidrug-resistant lineages, were associated with transmission clusters at multiple units, and these were collectively responsible for two-thirds of all infections. Carriage of carbapenemase genes (adjusted odds ratio, aOR = 2.08 [95% confidence interval, CI: 1.04, 4.14]; p = 0.04) and ESBL genes (aOR = 2.48 [95% CI: 1.26, 4.90]; p = 0.006) were significantly positively associated with transmission in a logistic regression model with site as a covariate. Limitations of this study include the lack of sufficient clinical data to allow high-resolution investigation of transmission dynamics and lack of facility-level data to investigate contributors to the observed differences in transmission burden across sites. Conclusions Nosocomial transmission contributes to a substantial proportion of K. pneumoniae sepsis in neonatal care units in Africa and South Asia. Reducing transmission within these settings through improved infection prevention and control and other measures could substantially reduce the neonatal sepsis burden. A high burden of transmission clusters is associated with the same drug-resistant lineages that are recognised as high-risk clones associated with hospital outbreaks in high-income countries, indicating global connectivity of the antimicrobial-resistant pathogen population.

阅读与讨论 → 访问原文 →

10.

arXiv (CS.CL) 2026-06-15 DOI: arXiv:2606.14626

Characterizing Cultural Localization in AI-Generated Stories

作者:

Shaily Bhatt ↗Supriti Vijay ↗Jeremiah Milbauer ↗Fernando Diaz ↗

The global use of artificial intelligence has increased interest in assessing the ability to generate culturally localized content, including stories. Cultural localization in stories often occurs through either templated localization – the use of cultural markers (e.g., names, locations) in a generic narrative – or holistic localization – the variation of plots, values, and themes, in addition to cultural markers. We propose a method to measure the degree to which content was generated through templated localization. Specifically, we identify the lexical tokens that distinguish stories across nationalities and measure the similarity of the narratives that remain after removing them. In stories generated by five models on 125 topics for 193 nationalities, our method is able to detect that only a small subset (9-17%) of the vocabulary accounts for the variation across nationalities and that the narratives that remain after removing them contain repeated multi-word sequences, suggesting the presence of a shared culturally-agnostic narrative template. Finally, we characterize the cultural markers for their stereotypicality and offensiveness, finding that markers from 19 countries, mostly located in the Global South, are on average offensive.

阅读与讨论 → 访问原文 →

11.

bioRxiv (Bioinfo) 2026-06-19 DOI: HASH:5f0a834997952017071bcad401de397e

SteerAF: Distogram-based Steering of AlphaFold2 toward Alternative Conformations

作者:

Tang ↗Zhu ↗Yang ↗Song ↗

End-to-end structure predictors, such as AlphaFold2, typically output only the dominant conformational state of a given protein, which is biased by the training data set. Existing strategies for recovering alternative conformations are often computationally expensive and offer limited biological interpretability. Here, we present SteerAF, an inference-time optimization framework based on AlphaFold2 that leverages information encoded in the distogram derived from deep multiple sequence alignments (MSAs) to predict alternative protein conformations. Across four benchmark datasets, SteerAF matches or surpasses existing methods in predicting alternative conformations for the majority of systems. Sparse MSA-feature modifications generated via block gradient ascent exhibit a strong correlation with experimentally characterized functional residues, recovering them with approximately 50% precision in the tested proteins. Furthermore, SteerAF enables effective decoy selection in the absence of experimental structures, and its predictions can serve as seed structures for molecular dynamics simulations to map conformational landscapes. Thus, SteerAF provides an efficient and interpretable approach for predicting alternative conformations, offering a framework that can be extended to other similar predictors and problems.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.LG) 2026-06-19 DOI: arXiv:2606.20326

Quantum-classical physics-informed Kolmogorov-Arnold networks for PDEs

作者:

Xiang Rao ↗Yuxuan Shen ↗

arXiv:2606.20326v1 Announce Type: new Abstract: We develop QCPIKAN, the first quantum-classical physics-informed Kolmogorov-Arnold network designed to solve partial differential equations (PDEs). Built upon Chebyshev-polynomial KAN layers and parameterized quantum circuits, this hybrid framework embeds physical constraints into the training loss to enforce physical consistency. Our theoretical investigations grounded in approximation theory prove that this design accelerates high-frequency error convergence to an exponential rate and effectively mitigates numerical dispersion. We validate the framework across three typical seepage scenarios in porous media, including single-phase flow, component transport and two-phase flow. Compared with existing quantum-classical physics-informed neural networks, QCPIKAN achieves superior performance in global prediction accuracy, local error control, dynamic evolution tracking and displacement front localization. This work provides a robust and efficient alternative for solving complex PDEs.

阅读与讨论 → 访问原文 →

13.

arXiv (math.PR) 2026-06-16 DOI: arXiv:2606.15089

A Machine-Checked Itô Calculus for Brownian Motion

作者:

Raphael Coelho ↗

arXiv:2606.15089v1 Announce Type: cross Abstract: We present a machine-checked development of the $L^2$ Itô calculus of Brownian motion on a bounded time interval $[0,T]$, formalized in Lean 4 on top of Mathlib and the BrownianMotion package. The development contains: the construction of the Itô integral as an isometry of Hilbert spaces, from a predictable-rectangle $\pi$-system through the density of simple adapted processes; the Itô integral as a process, proved to be an $L^2$-continuous martingale through a single structural identity (the integral at time $t$ is the conditional-expectation projection of its terminal value onto $\mathcal{F}t$), from which adaptedness, the martingale property, the contraction bound, and both the terminal and the time-indexed Itô isometries follow as corollaries; and Itô's formula for $C^3$ functions with bounded derivatives, including its time-dependent form $df = f_x,dB + (f_t + \tfrac12 f{xx}),dt$, obtained by a discrete-to-continuous argument through weighted quadratic variation and explicit $L^2$ remainder bounds. To our knowledge this includes the first machine-checked proof of Itô's formula, and the first machine-checked construction of the Itô integral as a martingale-valued process, in any proof assistant. We are deliberate about the boundary: the theory is the $L^2$ theory on $[0,T]$ with bounded-derivative integrand classes; localization to the unrestricted $C^2$ formula, integrators beyond Brownian motion, and pathwise statements are out of scope, and we say precisely why and where. The development is roughly 7,200 lines of Lean across 22 modules; every theorem is sorry-free, the axioms of each headline result are pinned to Mathlib's classical defaults by a build-enforced gate, and the whole is reproducible from a pinned toolchain.

阅读与讨论 → 访问原文 →

14.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.16074

PVminerLLM2: Improving Structured Extraction of Patient Voice via Preference Optimization

作者:

Samah Fodeh ↗Linhai Ma ↗Ganesh Puthiaraju ↗Srivani Talakokkul ↗Afshan Khan ↗Elyas Irankhah ↗Sreeraj Ramachandran ↗Ashley Hagaman ↗Sarah Lowe ↗Aimee Roundtree ↗

Motivation: Patient-generated text contains critical information on patients' lived experiences, social context, and care engagement, but remains largely unstructured, limiting its use in patient-centered outcomes research. Prior work introduced the PV-Miner benchmark and PVMinerLLM models for structured extraction. However, supervised fine-tuning (SFT) alone struggles with rare, fine-grained, and unevenly distributed errors, particularly in token-critical structured outputs. Results: We present PVminerLLM2, an improved set of LLMs for structured patient voice extraction that applies preference optimization to address token-critical errors beyond the reach of supervised fine-tuning. Our method introduces (i) a preference objective with token-level gated stabilization term that prevents degradation of absolute token likelihood under preference optimization, and (ii) confusion-aware preference pair construction to better capture low-separation distinctions. We further incorporate token-importance weighting and inverse-frequency reweighing to address token imbalance and class skew. Across multiple model sizes, PVMinerLLM2 consistently outperforms strong baselines, achieving gains of up to 4.43% (Code), 3.50% (Sub-code), and 1.55% (Span), and outperforms baseline LLM trained with existing preference optimization methods. Availability and Implementation: The supplementary material, code, evaluation scripts, and trained models for PVminerLLM2 are publicly available at: https://github.com/Data-Mining-Lab-Yale/PVminerLLM2

阅读与讨论 → 访问原文 →

15.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2606.07082

On the Geometry of On-Policy Distillation

作者:

Zhennan Shen ↗Yanshu Li ↗Qingyu Yin ↗Chak Tou Leong ↗Zhilin Wang ↗Yanxu Chen ↗Rongduo Han ↗Sunbowen Lee ↗Yi R. Fung ↗

arXiv:2606.07082v2 Announce Type: replace-cross Abstract: On-policy distillation (OPD) is increasingly used to improve large language model reasoning, but its training dynamics remain poorly understood. We characterize the trajectory of OPD updates in parameter space and compare it with supervised fine-tuning (SFT) and reinforcement learning with verifiable rewards (RLVR). A suite of parameter-space diagnostics consistently places OPD in a relaxed off-principal regime: compared with SFT, its updates affect fewer weights and avoid principal directions more strongly, while compared with RLVR, they remain less tightly constrained. Beyond this static localization, OPD exhibits subspace locking: its cumulative updates rapidly enter a narrow low-dimensional channel. Constraining training to the update subspace formed early in training preserves OPD performance but substantially degrades SFT, indicating that the locked subspace is functionally sufficient for OPD. Control experiments further show that sparsifying the update tokens and shifting rollout generation off-policy preserve the rank dynamics, whereas mixing the OPD objective with RLVR changes them. Overall, these results suggest that OPD is not merely an intermediate point between SFT and RLVR, but induces its own update geometry in parameter space.

阅读与讨论 → 访问原文 →

16.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.13111

M\"OVE: A Holistic LLM Benchmark for the German Public Sector

作者:

Camilla Dalerci ↗Thilo Michael ↗Robin Schaefer ↗Daniel Weinland ↗

We present M\"OVE (Modelle für die \"Offentliche Verwaltung Evaluieren), a holistic benchmark for evaluating large language models (LLMs) in the context of the German public sector. While LLMs are increasingly adopted in public administration, model selection remains largely ad hoc, and existing benchmarks offer limited guidance: they are predominantly English-centric, US-centric in content, and focus exclusively on task performance. M\"OVE addresses these gaps by evaluating 39 models across two complementary dimensions. Performance criteria cover summarization, question answering, and topic extraction. Governance criteria assess hallucination tendencies, energy consumption, provider transparency, and alignment with German constitutional values and knowledge about positions by German political parties. In total, we utilize ten German-language datasets, including gold- and silverstandard datasets that we constructed to reflect public-administration domains. We employ a multi-metric evaluation strategy combining classical NLP metrics, embedding-based methods, and LLM-as-a-judge approaches. Our results show that no single model dominates across all criteria: top performers differ between tasks, and model size alone is a poor predictor of quality. We further evaluate the benchmark itself, analyzing its statistical precision, LLM judge reliability, the impact of our private datasets on model rankings, the sensitivity of our results to prompt formulation, and the validity of our energy consumption estimates. M\"OVE is designed as a living benchmark under active development; results are publicly available at https://moeve.bundesdruckerei.de/.

阅读与讨论 → 访问原文 →

17.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.11692

Evaluation of Alternative-Based Information Systems for Deliberative Polling using an Agentic Simulator

作者:

Rwaida Alssadi ↗Khulud Alawaji ↗Balaji Kasula ↗Muntaser Syed ↗Badria Alfurhood ↗Markus Zanker ↗Marius Silaghi ↗

arXiv:2606.11692v1 Announce Type: cross Abstract: Deliberative polling promises to improve collective decision-making by exposing shareholders to a broad range of arguments before they vote. Yet ensuring that every voter encounters a representative sample of the reason space, the coverage problem, remains an open challenge, particularly at scale and in adversarial or strategically motivated electorates. This paper introduces a way of evaluating solutions using the LLM-based Agentic Bipolar Argumentation Simulator, grounded in a framework which formalises a poll as a six-tuple of endorsing and opposing justifications, attack and enhance relations, and shareholder- and relation-weights. ABAS simulates N autonomous shareholder agents, each assigned a latent opinion according to desired distributions in [-1, 1], who sequentially vote, choose or author justifications, and optionally submit argumentation-graph links. The simulator implements recommendations that rank existing justifications by their observable endorsement mass. It evaluates the mechanism's success by coverage, namely the fraction of the corpus reason-tag set represented in the K recommendations presented to each shareholder, as a solution to the NP-hard Subsuming Justification Problem. Reported experiments characterise how creativity rate (pown), recommendation size (K), argumentation density (plinks), and population size (N) affect coverage and corpus diversity. In an authenticated electorate where Sybil attacks are impossible and only the relation graph is gameable, we stress-test the scoring with coordinated strategic voting attacks: a tag-flood attack collapses coverage, while author-count relation weighting through a reversed-PageRank rule resists the flood markedly better than uniform weights.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2606.11260

RAIL: Rethinking Auditory Intelligence in Large Audio-Language Models with a CHC-Grounded Benchmark

作者:

Hongyu Jin ↗Siyi Wang ↗Yang Xiao ↗Jiaheng Dong ↗Shihong Tan ↗Kaiyuan peng ↗Georgiana Juravle ↗Shanquan Chen ↗Gongping Huang ↗Hong Jia ↗Eun-Jung Holden ↗James Bailey ↗…

arXiv:2606.11260v1 Announce Type: cross Abstract: Humans process rich auditory environments through tightly integrated cognitive capabilities such as audio perception, audio reasoning, and memory. Despite recent progress in large audio-language models (LALMs) across speech understanding and multimodal audio reasoning, current evaluation paradigms remain largely task- or modality-centric, focusing on end performance while overlooking underlying auditory cognitive behaviours. This reveals a fundamental gap between how auditory cognition is understood in humans and how it is evaluated in LALMs, particularly in the lack of frameworks that operationalise cognitive principles beyond task-level metrics to systematically capture model behaviour. In this work, we introduce RAIL, a human-centric evaluation paradigm grounded in the Cattell-Horn-Carroll (CHC) cognitive framework. RAIL formalises auditory cognition into five core capabilities and develop them into structured evaluation tasks that probe how models process, retain, and integrate auditory information. We further construct a cognitively grounded benchmark with principled data curation and human-aligned evaluation protocols. Evaluating 26 state-of-the-art LALMs, we find that current models exhibit highly uneven performance across cognitive abilities. RAIL establishes a new evaluation paradigm that moves beyond task-centric benchmarking toward cognitively grounded assessment of auditory intelligence.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.CL) 2026-06-19 DOI: arXiv:2606.20225

Actionable Activation Directions for Detecting and Mitigating Emergent Misalignment Across Language Model Families

作者:

Abdul Rafay Syed ↗

Fine-tuning language models on insecure code induces emergent misalignment with poorly understood internal structure. We investigate whether this misalignment corresponds to a causally actionable activation-space direction shared across architectures. Across four instruction-tuned model families (Qwen2.5-1.5B, Gemma-2-2B, Llama-3.2-1B, Ministral-3-3B) finetuned identically, a difference-in-means direction achieves 99.6% separation of aligned and misaligned activations at each model's final layer. Causal steering by subtracting this direction reduces code spillover by 21-51 points, while a secure-code control confirms content specificity. Cross-architecture transfer via ridge regression maps yields large behavioral suppression (up to 46 points) but fails specificity controls as random and orthogonal directions perform comparably. We identify a two-tier specificity structure: within-model directions are causally specific and actionable; cross-model directions are causally real but non-specific. An asymmetric transfer topology emerges, with Gemma and Qwen acting as geometric donors and Llama as a receiver. These findings define the limits of linear cross-architecture correction and recommend within-model probing for auditing.

阅读与讨论 → 访问原文 →

20.

arXiv (CS.AI) 2026-06-24 DOI: arXiv:2606.03070

ASymPO: Asymmetric-Scale Policy Optimization for Asynchronous LLM Post-Training Without Behavior Information

作者:

Zehua Liu ↗Yuxuan Yao ↗Xiaojin Fu ↗Tao Zhong ↗Mingxuan Yuan ↗

arXiv:2606.03070v3 Announce Type: replace-cross Abstract: Asynchronous reinforcement learning can improve language-model post-training throughput by decoupling response generation from policy optimization, but stale responses introduce distribution drift. Standard behavior-corrected methods control this drift with behavior-policy probabilities, importance ratios, or clipping, which requires token-aligned, versioned, and numerically consistent behavior log-probabilities across rollout and learner systems. We ask whether asynchronous group-relative RL can instead be stabilized using only current-policy probabilities. We identify a scale-imbalance failure mode: when stale responses are evaluated under the current policy, positive and negative loss terms can appear at different negative log-probability scales, so zero-sum advantages no longer imply balanced loss contributions. We propose Asymmetric-Scale Policy Optimization (ASymPO), which normalizes each response's token loss by its current average token negative log-probability. ASymPO requires no behavior-policy probabilities, restores response-level zero-sum balance, and preserves a nonzero learning signal. We also introduce Scaled Policy Optimization (SPO), a fixed negative-scaling baseline, and evaluate both current-policy-only objectives in asynchronous mathematical reasoning post-training.

阅读与讨论 → 访问原文 →

21.

arXiv (quant-ph) 2026-06-24 DOI: arXiv:2602.15800

Entanglement in the Dicke subspace

作者:

Aabhas Gulati ↗Ion Nechita ↗Cl\'ement Pellegrini ↗

arXiv:2602.15800v2 Announce Type: replace Abstract: We provide a complete mathematical theory for the entanglement of mixtures of Dicke states. These quantum states form an important subclass of bosonic states arising in the study of indistinguishable particles. We introduce a tensor-based parametrization where the diagonal entries of these states are encoded as a symmetric tensor, enabling a direct translation between entanglement properties and well-studied convex cones of tensors. Our results bridge multipartite entanglement theory with semialgebraic geometry and the theory of completely positive and copositive tensors. This dictionary maps separability to completely positive tensors, the PPT property to moment tensors, entanglement witnesses to copositive tensors, and decomposable witnesses to sum of squares tensors. Using this framework, we construct explicit PPT entangled states in three or more qutrits, disproving a recent conjecture. We establish that PPT entanglement exists for all multipartite systems with local dimension d >= 3 and n >= 3 parties. We also show that, for mixtures of Dicke states, the PPT condition with respect to the most balanced bipartition implies all other PPT conditions. We further connect bosonic extendibility of mixtures of Dicke states to the duals of known hierarchies for non-negative polynomials, such as the ones by Reznick and Polya. We thus provide semidefinite programming relaxations for separability and entanglement testing in the Dicke subspace.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2606.11976

Exploration Structure in LLM Agents for Multi-File Change Localization

作者:

Akeela Darryl Fattha ↗Kia Ying Chua ↗Lingxiao Jiang ↗Laura Wynter ↗

arXiv:2606.11976v1 Announce Type: cross Abstract: Software engineering tools increasingly rely on LLM based agents to localize files to change to resolve a software issue. Most AI agents explore repositories linearly, that is, visiting one directory or file per step. We postulate that this is a structural mismatch for changes that span several subsystems. We compare linear sequential exploration against non-linear, domain-scoped parallel agentic exploration. Using SWE Bench Pro as initial benchmark, we focus on ansible as an exemplar. We construct an approach for persistent-session evaluation of GitHub issues anchored at a single base commit. We compare our non-linear domain-agent file traversal system against a base LLM without direct repository access, a single agent Recursive Language Model (RLM) baseline with a persistent Python REPL and an external CLI baseline using Codex 5.5 High. Domain scoped parallel agent spawning with a small Haiku-class model achieves the highest micro F1 among Haiku class models by a large margin. Domain-agents is the second highest behind only the much larger Codex 5.5 High on our own expanded benchmark including over more recent PRs from 2025 and 2026. On the original, curated, 2020 SWE-bench Pro benchmark, a larger Sonnet plain LLM baseline attains higher micro F1 by predicting few files, leading to higher precision, but at significantly lower all gold recall. We also present three additional findings. First, documentation evolution is a latent dependency unresolved by any approach. Second, naive file system access can degrade localization driven by test-file over prediction. Lastly, forced multi-agent consultation does not measurably help and raises token cost substantially.

阅读与讨论 → 访问原文 →

23.

arXiv (CS.LG) 2026-06-15 DOI: arXiv:2606.14636

Graph Diffusion Residuals for Control-Function Instrumental Variables

作者:

Rui Wu ↗Zongyuan Chen ↗Hong Xie ↗Defu Lian ↗Enhong Chen ↗

arXiv:2606.14636v1 Announce Type: new Abstract: Control-function instrumental variable estimators need a first-stage residual, not merely a first-stage prediction. High-capacity first stages can interpolate treatment and leave too little residual information for the outcome equation. We study Adaptive Anisotropic Instrumental Heat Flow (A-IHF), a deterministic graph-diffusion residual extractor for flexible control functions. A-IHF treats treatment as a signal on a graph of first-stage features, uses pilot diffusion to detect large treatment jumps, attenuates conductance across those jumps, and computes the generated control with a sparse graph resolvent. Its observational selection rule uses only $(Z,X)$, combining graph generalized cross-validation, roughness, residualized-treatment relevance, and graph-admissibility filtering. The analysis decomposes error into structural leakage, residual attenuation, and residualized treatment variation, yielding finite-sample bounds, graph-admissibility rates under latent piecewise-smooth geometry, and finite-path selection calibration. Across 54 synthetic benchmark cells with tuned graph, kernel, tree, boosting, series, and neural control-function baselines, guarded observational A-IHF has the lowest average structural-response MSE; the A-IHF family beats the best non-A-IHF baseline in 32 cells. Performance is strongest when the graph captures piecewise-smooth first-stage structure.

阅读与讨论 → 访问原文 →

24.

arXiv (quant-ph) 2026-06-24 DOI: arXiv:2506.13724

Logical qubits with erasure conversion using metastable neutral atoms

作者:

Bichen Zhang ↗Genyue Liu ↗Guillaume Bornet ↗Sebastian P. Horvath ↗Pai Peng ↗Shuo Ma ↗Shilin Huang ↗Shruti Puri ↗Jeff D. Thompson ↗

arXiv:2506.13724v2 Announce Type: replace Abstract: Implementing large-scale quantum algorithms with practical advantage will require fault-tolerance achieved through quantum error correction, but the associated overhead is prohibitive. This overhead can be reduced by engineering physical qubits with fewer errors, and by shaping the residual errors to be more easily correctable. In this work, we demonstrate quantum error correcting codes and logical qubit circuits in a metastable ytterbium-171 nuclear spin qubit with a noise bias towards erasure errors. These errors can be located separately from any syndrome information diagnosing the error, and we demonstrate adaptive circuit execution based on erasure information. We show that dephasing errors on the qubit during coherent transport can be strongly suppressed, and implement entangling gates that maintain a high fidelity in the presence of gate beam inhomogeneity or pointing errors. Furthermore, we demonstrate logical qubit encoding in the [[4, 2, 2]] code, with error correction during decoding based on mid-circuit erasure measurements despite the fact that the code is too small to correct any Pauli errors. Finally, we demonstrate logical qubit teleportation between multiple code blocks with conditionally selected ancillas based on mid-circuit erasure checks, a key part of leakage-robust error correction schemes using neutral atoms.

阅读与讨论 → 访问原文 →

25.

medRxiv (Medicine) 2026-06-10 DOI: HASH:2dc5caff0b1a9ca145dbb64de687eee5

Exploratory Assessment of Pulsed-Wave Doppler Representations of Lung Sounds Using Deep Learning: An In-Vitro Phantom Study

作者:

Saad ↗A. A ↗Murthi ↗S. B ↗Boctor ↗E. M ↗Teeter ↗W. A ↗Seam ↗

The increasing availability of portable ultrasound systems motivates exploration of novel approaches to respiratory signal assessment. In this in-vitro study, we investigate whether pulsed-wave (PW) Doppler ultrasound can capture structured spectral patterns from replayed lung sound recordings. Digitized respiratory sounds were replayed through a tissue-mimicking ultrasound phantom, generating 1,478 PW Doppler spectral images from recordings associated with healthy subjects and several externally labeled disease categories. Exploratory classification experiments using a ResNet-18 architecture demonstrated that these Doppler representations contain learnable differences under controlled conditions. These findings motivate further investigation into PW Doppler as a potential representation of respiratory acoustics.

阅读与讨论 → 访问原文 →

探索全球前沿学术脉络