论文广场 - AcademicHub

01.

arXiv (quant-ph) 2026-06-16 DOI: arXiv:2512.06718

Witnessing Spin-Orbital Entanglement using Resonant Inelastic X-Ray Scattering

作者:

Zecheng Shen ↗Shuhan Ding ↗Zijun Zhao ↗Francesco A. Evangelista ↗Yao Wang ↗

arXiv:2512.06718v2 Announce Type: replace Abstract: Entanglement plays a central role in quantum technologies, yet its characterization and control in materials remain challenging. Recent developments in spectrum-based entanglement witnesses have enabled new strategies for quantifying many-body entanglement in macroscopic materials. Here, we develop a protocol for detecting spin-orbital entanglement using experiment-accessible resonant inelastic x-ray scattering (RIXS). Central to our approach is the construction of a Hermitian generator from experimentally measurable spectra, which allows us to compute the quantum Fisher information (QFI) available in spin–orbital systems. The resulting QFI provides upper bounds for $k$-producible states and thus serves as a robust witness of spin-orbital entanglement. To account for realistic experimental limitations, we further extend our framework to include relaxed QFI bounds applicable to measurements lacking full polarization resolution.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.16364

Looking Is Not Picking: An Attention-Segment Account of Tool-Selection Failures in LLM Agents

作者:

Shiyang Chen ↗

arXiv:2606.16364v1 Announce Type: new Abstract: LLM agents mis-call tools, and the natural guess is that the model failed to see the right tool in a crowded harness. We show the opposite through a lens concurrent work sets aside – the model's attention to labeled tool-definition segments. On real BFCL failures, by per-candidate attention argmax the model attends most to the correct tool 80% of the time (vs. 21% chance), and the gold is the under-attended segment on only 10%: it looks at the right tool and still picks wrong. This directly refutes the intuitive "crowded-harness / lost-in-the-middle" explanation: the failure is at the decision readout, not the harness, and we pin it there three ways. (1) Input vs. readout: repairing the prompt (reordering or duplicating the gold tool) recovers

阅读与讨论 → 访问原文 →

03.

arXiv (CS.CL) 2026-06-24 DOI: arXiv:2606.24172

A P\={a}ninian Foundation for Indic Language Processing

作者:

Ritwik Banerjee ↗Lav R. Varshney ↗

More than a billion people communicate in Indic languages, yet the natural language processing infrastructure serving them remains fragmented and underdeveloped. The cause is structural: the field organizes its tools and benchmarks around individual languages or small subsets of genealogical language families, building separate analyzers, parsers, and datasets for each language and starting over for the next. This overlooks a deep regularity. Through more than two millennia of convergence around Sanskrit, Indic languages came to share a morphosyntactic architecture formalized in P\={a}nini's grammar, the Ast\={a}dhy\={a}y\={i}. This cuts across genealogical lines, uniting languages through a common framework. We argue that this P\={a}ninian framework supplies a unifying computational architecture the field has lacked, and that benchmarks grounded explicitly in it would make Indic language systems more accurate, more data-efficient, and more transferable, effectively merging many apparently disparate and sparse Indic language resources into a single high-resource metalanguage bedrock. We propose a four-part benchmark suite to render this shared architecture explicit, measurable, and ready to be leveraged for practical applications. Moreover, we underscore the question it raises for interpretability research: whether neural models trained on these languages come to represent P\={a}nini's categories on their own.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.16190

Embedded Arena: Iterative Optimization via Hardware Feedback

作者:

arXiv:2606.16190v1 Announce Type: cross Abstract: Embedded devices from wildlife monitoring stations to clinical wearables require local AI inference due to latency, communication, or privacy constraints. Optimizing models for heterogeneous microcontrollers (MCUs) requires simultaneously satisfying hard physical constraints on memory, power, and temperature while preserving accuracy, a multidimensional optimization that is today performed manually by experts. We ask whether an LLM agent can autonomously navigate this complex, multi-turn pipeline guided by real hardware feedback, and introduce a hardware-in-the-loop agent arena in which the agent iteratively refines both model and firmware – compiling, flashing, and measuring on real hardware – to enable closed-loop optimization. Frontier models, including Claude Opus 4.7 and Gemini 3.1 Pro, fail entirely without hardware feedback (0% deployment success), whereas our hardware-in-the-loop formulation achieves the first successful deployment within three iterations and can surpass human expert results within seven. This agentic co-optimization achieves 250x compression for vision models with

阅读与讨论 → 访问原文 →

05.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2606.11195

From Consumption to Reflection: Designing Human-AI Relations for Stable Reasoning

作者:

Rikard Rosenbacke ↗Carl Rosenbacke ↗Victor Rosenbacke ↗Martin McKee ↗

arXiv:2606.11195v1 Announce Type: cross Abstract: Large language models (LLMs) have transformed how humans access information, but not how we reason with it. Their fluency accelerates consumption while bypassing the slow, reflective processes that underpin sound judgment. This paper introduces Relational Reflective Intelligence (RRI), an inference-time governance layer that operationalizes reflection through auditable reasoning loops. RRI operates not inside the model but around it, providing a practical structure for stable, auditable reasoning between humans and LLMs. The core premise is that LLMs inherit cognitive vulnerabilities similar to those that shape human thought: reliance on intuitive shortcuts, confusion between representation and reality, and a preference for coherence over falsification. When humans and models share these tendencies, their errors compound. We refer to this as relational drift, a failure that arises from interaction rather than from the model alone. Addressing this requires a shift from modeling relations between words to structuring relations between model outputs and human reasoning. RRI provides this missing layer through three components: the Rose-Frame, which identifies likely breakdowns in reasoning; the Architect's Pen, which introduces targeted reflection steps at critical moments; and an inference-time workflow that embeds these steps without retraining the model. Together, these elements transform human-AI interaction into a joint reasoning system with explicit checkpoints, conflict surfacing, and an auditable trail of assumptions. Rather than making machines think like humans or forcing humans to reason like machines, RRI creates a structured interaction in which both compensate for each other's limitations. It reframes AI safety as a cognitive architecture problem, where reliable decisions depend on embedding reflection directly into the interaction process.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.15899

SkillVetBench: LLM-as-Judge for Multi-Dimensional Security Risk Evaluation in Open-Source LLM Agent Skills

作者:

Ismail Hossain ↗Sai Puppala ↗Md Jahangir Alam ↗Tanzim Ahad ↗Sajedul Talukder ↗

arXiv:2606.15899v1 Announce Type: cross Abstract: Open-source LLM agent ecosystems are growing rapidly, yet the security of community-contributed skills - modular tool definitions that extend agent capabilities - remains largely unvetted. The gap we fill: existing scanners operate at the code layer and are structurally blind to instruction-layer and multi-agent risk - natural-language directives that hijack an agent, exfiltrate data through encoded side channels, or chain harm across pipelines - so what is needed is a semantic, multi-dimensional vetting system rather than another signature matcher. We present SKILLVETBENCH, a live public leaderboard on Hugging Face that uses an LLM-as-Judge to vet agent skills. What is new: SARS (Skill Agentic Risk Score), a five-dimensional agentic-risk metric with a principled weighted formula for instruction-following systems. What is integrated: full CVSS v4.0 vector decomposition and a ClawHub dual-view that places our LLM-generated review beside the official marketplace verdict. What is demonstrated: drawing on our companion benchmark paper [ 1], the LLM-as-Judge stage achieves zero false negatives across 78 confirmed-malicious skills and zero false positives across 22 benign controls, while the best static baseline (SKILLSIEVE) still misses 15%; for instruction-layer categories such as Prompt Injection and Memory Poisoning, conventional tools miss between 89% and 100% of threats (e.g., CODEBERT detects none of nine memory-poisoning skills). Detection rates vary from 35% to 95% across four LLM evaluators, motivating ensemble scoring in production deployments.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2601.14968

InstructTime++: Time Series Classification with Multimodal Language Modeling via Implicit Feature Enhancement

作者:

Mingyue Cheng ↗Xiaoyu Tao ↗Huajian Zhang ↗Qi Liu ↗Zhiding Liu ↗Yucong Luo ↗Yiheng Chen ↗Enhong Chen ↗

arXiv:2601.14968v2 Announce Type: replace-cross Abstract: Most existing time series classification methods adopt a discriminative paradigm that maps input sequences directly to one-hot encoded class labels. While effective, this paradigm struggles to incorporate contextual features and fails to capture semantic relationships among classes. To address these limitations, we propose InstructTime, a novel framework that reformulates time series classification as a multimodal generative task. Specifically, continuous numerical sequences, contextual textual features, and task instructions are treated as multimodal inputs, while class labels are generated as textual outputs by tuned language models. To bridge the modality gap, InstructTime introduces a time series discretization module that converts continuous sequences into discrete temporal tokens, together with an alignment projection layer and a generative self-supervised pre-training strategy to enhance cross-modal representation alignment. Building upon this framework, we further propose InstructTime++, which extends InstructTime by incorporating implicit feature modeling to compensate for the limited inductive bias of language models. InstructTime++ leverages specialized toolkits to mine informative implicit patterns from raw time series and contextual inputs, including statistical feature extraction and vision-language-based image captioning, and translates them into textual descriptions for seamless integration. Extensive experiments on multiple benchmark datasets demonstrate the superior performance of InstructTime++.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2606.14737

Learning Topological Representations for Molecular Dynamics

作者:

Dominik Geng ↗Florian Graf ↗Martin Uray ↗Roland Kwitt ↗

arXiv:2606.14737v1 Announce Type: cross Abstract: Molecular dynamics (MD) simulations generate trajectories in a high-dimensional configuration space whose analysis critically depends on molecular descriptors, typically handcrafted observables or learned kinetic embeddings. Designing descriptors that are both expressive and broadly applicable, however, remains challenging. We study persistent homology (PH) as a general-purpose representation for MD and introduce the masked Flood complex, a protein-tailored modification of a recently introduced simplicial complex construction that emphasizes inter-residue structure at low computational cost. Vectorized persistence diagrams then provide information-rich, geometry-aware summaries of protein conformations, which we evaluate on protein class prediction, frame-level observable regression, and Markov state model (MSM) estimation from learned low-dimensional coordinates in a single shared representation space. Results on the mdCATH dataset show that PH-based descriptors are competitive across tasks, with masked Flood PH yielding the most consistent overall performance. Further, when using topologically-informed MSMs as a drop-in replacement within the recent MarS-FM framework for generative modeling of protein conformations, we obtain consistently better ensemble statistics than MSMs based on physical observables. Finally, we explore the transferability of the generative model to qualitatively different, fast folding, proteins.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2606.11218

An Ethical eValuation Agent (EeVA): Results of a Proof-of-Concept Test on a Prototype Agentic-like Workflow to Assist Ethical Deliberations

作者:

Stephen Milford ↗B. Zara Malgir ↗Miguel Vazquez ↗

arXiv:2606.11218v1 Announce Type: cross Abstract: Ethical deliberation is often misunderstood as a search for single right or wrong answers, creating difficulties for non-ethically trained personnel who must address ethically laden challenges. We developed EeVA, an agentic-like LLM-based workflow designed to support comparative ethical reflection rather than deliver definitive ethical answers. EeVA was programmed in n8n using three interconnected workflows: starter, worker, and emitter. It evaluated uploaded use cases against 10 ethical frameworks through evaluator and synthesis prompts. Proof-of-concept testing used three published cases from urban mobility, peer-to-peer energy trading, and social-service resource allocation. Across all cases, EeVA produced consistently structured framework-specific evaluations and integrated syntheses. Outputs differentiated between frameworks, identified convergences and divergences, recommended modifications to increase alignment, and highlighted persistent ethical tensions. Syntheses were readable for non-specialists and shifted attention away from simplistic answers toward design conditions, safeguards, and areas where full cross-framework agreement was unlikely. The findings suggest that LLMs can be organised into usable workflows that preserve ethical plurality while helping bridge the communicative gap between ethicists and non-ethically trained personnel. EeVA's value lies not in replacing ethicists or resolving moral disagreement, but in scaffolding structured ethical deliberation. EeVA offers a promising proof of concept for supporting ethical reflection where access to ethics expertise is limited. Further work is needed on reproducibility, human evaluation, user testing, and efficiency before it can be considered a mature tool.

阅读与讨论 → 访问原文 →

10.

arXiv (quant-ph) 2026-06-16 DOI: arXiv:2606.15113

Optimal Toffoli-Depth Multi-Controlled Toffoli Decomposition in 2D Qubit Layout

作者:

Anik Basu Bhaumik ↗Suman Dutta ↗Anupam Chattopadhyay ↗

arXiv:2606.15113v1 Announce Type: new Abstract: The multi-controlled Toffoli (MCT) gate is a key primitive in quantum arithmetic, oracle construction, and quantum cryptanalysis. Although recent work has established optimal Toffoli-depth MCT decompositions under all-to-all qubit connectivity, their realization on near-term quantum hardware with restricted qubit connectivity remains largely unexplored. While general-purpose quantum mappers can route arbitrary circuits, they do not explicitly exploit the repeated interaction patterns inherent in MCT decompositions. In our present paper, we study architecture-aware mappings of optimal Toffoli-depth MCT decompositions onto restricted two-dimensional qubit layouts. We begin with a structured geometric placements that preserve the parallelism of state-of-the-art Toffoli and MCT decompositions with no additional depth overhead. We further introduce a motif-based packing framework in which decomposition layers are represented by interaction motifs derived from basic Toffoli gates. By embedding these motifs vertex-disjointly into hardware graphs, we characterize the minimum-size topologies supporting the required qubit resources and derive explicit bounds on the resulting depth overhead under tight qubit budgets. Finally, we compare these bounds with routing-aware placement heuristics and empirically evaluate the effectiveness of embedding different motifs across a range of hardware topologies.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.CL) 2026-06-19 DOI: arXiv:2606.04075

Large Language Models Hack Rewards, and Society

作者:

Wei Liu ↗Xinyi Mou ↗Hanqi Yan ↗Zhongyu Wei ↗Yulan He ↗

Reinforcement learning (RL) has become a dominant post-training paradigm, enabling large language models (LLMs) to learn from rewards. We observe that societal regulations are structurally similar to reward functions. They define measurable outcomes, thresholds, and exceptions, while often leaving institutional intent only partially specified. We hypothesise that the RL training process may exploit these gaps and therefore ask whether models' well-known tendency to hack reward functions during RL can scale into a more consequential failure mode named societal hacking: discovering loopholes in the rules society runs on. To study this phenomenon, we introduce SocioHack, a sandbox of 72 societal environments, and find that within these environments, reward hacking naturally emerges and leads to regulatory loophole discovery. Models learn to hack the social rules and generate strategies that remain technically compliant while defeating regulatory intent, and current LLM safeguards provide only limited mitigation. Therefore, collecting in-the-wild feedback for model training requires greater caution, and we need a next-generation post-training paradigm for safely iterating LLMs in real society.=

阅读与讨论 → 访问原文 →

12.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.16541

The Faithfulness Gap: Certifying Semantic Equivalence Between Natural-Language and Formal Mathematical Statements

作者:

Noor Islam S. Mohammad ↗Tamim Sheikh ↗

arXiv:2606.16541v1 Announce Type: new Abstract: Autoformalization, translating natural-language mathematics into formal proof assistants, is bottlenecked not by translation fluency but by faithfulness: a formal statement can typecheck and be provable, yet still encode a different theorem than the source intended. We introduce Bidirectional Provability Fingerprinting (\bpf{}), a framework that certifies faithfulness by characterizing each candidate through its forward and backward consequence neighborhoods in the ambient theory and matching these against probes derived from the natural-language statement. We further introduce four novel components: (i) Counterfactual Probe Generation (\cpg{}), a contrastive procedure that synthesizes probes targeting specific drift directions; (ii) the Equivalence Spectrum, a continuous faithfulness score that replaces brittle binary verdicts; (iii) Adaptive Probe Budget Allocation (\apba{}), an information-theoretic budget router; and (iv) Faithfulness-Guided Decoding (\fgd{}), which uses \bpf{} signals as a reward during autoformalization. We prove a drift detection theorem and a PAC-faithfulness result establishing that the equivalence class of a natural language statement is learnable from $\mathcal{O}(\log(1/\delta)/\varepsilon)$ probes under mild assumptions. We release \driftbench{}, a benchmark of $2{,}183$ NL/Lean~4 pairs with controlled drift labels across six subfields of mathlib4. \bpf{}\,+\,\cpg{} detects $89.6\%$ of drifted formalizations at a $3.0\%$ false-positive rate-against $41.2\%$ for typecheck and $63.3\%$ for LLM-judge baselines, and \fgd{} reduces the rate at which a state-of-the-art autoformalizer emits drifted statements by $47\%$. https://pmlrbd.github.io/BPF/

阅读与讨论 → 访问原文 →

13.

arXiv (CS.LG) 2026-06-17 DOI: arXiv:2606.18236

Sign-Rank, Index, and List Replicability: Connections and Separations

作者:

Ari Blondal ↗Hamed Hatami ↗Pooya Hatami ↗Chavdar Lalov ↗Sivan Tretiak ↗

arXiv:2606.18236v1 Announce Type: new Abstract: In learning theory, the sign rank of a binary concept class captures the smallest dimension in which it can be represented by points and halfspaces. Despite tremendous interest, lower bounds on sign rank are notoriously difficult to come by. Two recent approaches to the problem establish lower bounds on sign rank by measures that are easier to analyze: the $\mathbb{Z}_2$-index and the list replicability number. We order these measures, showing that the $\mathbb{Z}_2$-index is upper-bounded by a linear function of the list replicability number. As a main consequence, we obtain a strong separation between sign rank and $\mathbb{Z}_2$-index, thereby resolving a question of Frick, Hosseini, and Vasileuski. This motivates a thorough study of list replicability, the stronger of the two lower-bounding measures. We establish upper bounds on the list replicability number by two combinatorial measures: height and minimum star number. We also prove a fundamental composition result, showing that the product of two concept classes has list replicability number bounded by the sum of the list replicability numbers of the two classes.

阅读与讨论 → 访问原文 →

14.

arXiv (CS.AI) 2026-06-24 DOI: arXiv:2606.24115

A Benchmark for Hallucination Detection in VLMs for Gastrointestinal Endoscopy

作者:

Aminu Lawal ↗Niyoj Oli ↗Sachin Acharya ↗Prashnna Gyawali ↗Maria Carmen Romano ↗Binod Bhattarai ↗

arXiv:2606.24115v1 Announce Type: cross Abstract: Vision-language models (VLMs) are prone to hallucination, which remains a major barrier to their safe deployment in clinical practice. To date, most hallucination detection methods have been evaluated on radiology benchmarks such as MIMIC-CXR and VQA-RAD, while gastrointestinal (GI) endoscopy remains largely underexplored. In this paper, we benchmark nine hallucination detection methods on the Gut-VLM dataset, a GI diagnostic Visual Question Answering (VQA) dataset with 4,392 test VQA pairs, across five VLMs (MedGemma-4B, MedGemma-27B, LLaVA-Med-7B, LLaVA-v1.6-7B, and Lingshu-32B). The methods span three categories: black-box methods (RadFlag, SelfCheckGPT-NLI), gray-box methods (AvgProb, AvgEnt, MaxProb, MaxEnt, Semantic Entropy, and VASE), and a white-box method (ReXTrust). Our results show that ReXTrust, a white-box method, achieves the highest AUC across all five models, outperforming the strongest alternative method on each VLM by a statistically significant margin (paired permutation test, p < 0.001 in all cases), reaching a peak AUC of 93.0 on MedGemma-4B. White-box hidden-state access provides a consistent advantage of 19.5 AUC points on average (range: 9.5–33.5), with ReXTrust maintaining strong performance even on LLaVA-v1.6-7B (AUC 79.9), where black-box methods and clustering-based gray-box methods collapse to near-chance performance. Among non-white-box methods, token-level gray-box statistics (MaxEnt, MaxProb) are the strongest alternatives, outperforming both clustering-based gray-box methods (Semantic Entropy, VASE) and black-box approaches on average. We further identify confident confabulation, a failure mode in which models hallucinate with high inter-sample consistency or high token-level probability, as a systemic failure for both consistency and uncertainty-based methods.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.CL) 2026-06-19 DOI: arXiv:2606.19552

LaViSA: A Language and Vision Structural Ambiguity Benchmark

作者:

Lee Sangmyeong ↗Shun Inadumi ↗Koichiro Yoshino ↗

Structural ambiguity arises when a single sentence admits multiple valid interpretations due to its syntactic structure, posing a fundamental challenge for language understanding. Visual scenes serve as useful cues for resolving such ambiguity, and Vision and Language Models (VLMs) need to be capable of deriving possible semantic interpretations from visual scenes. We introduce Language and Vision Structural Ambiguity (LaViSA), a benchmark designed to evaluate the ability of VLMs to resolve structural ambiguity leveraging visual scenes. LaViSA consists of ambiguous sentences, their disambiguated sentences, and corresponding images of these disambiguated sentences across seven ambiguity categories. Using LaViSA, we conduct a comprehensive evaluation of diverse VLMs, including both proprietary and open-source models with varying parameter scales and reasoning capabilities. Experimental results show that although recent VLMs can leverage visual scenes to resolve structural ambiguity to a some extent, they still struggle with certain ambiguity types and visually subtle semantic distinctions, indicating remaining limitations in resolving structural ambiguity using visual scenes.

阅读与讨论 → 访问原文 →

16.

arXiv (quant-ph) 2026-06-17 DOI: arXiv:2606.17177

Hybrid Ferromagnet-SNSPDs: Single photon induced order-to-disorder transition in ferromagnets coupled to thin film superconductors

作者:

Leif Bauer ↗Daien He ↗Sathwik Bharadwaj ↗Zubin Jacob ↗

arXiv:2606.17177v1 Announce Type: cross Abstract: The development of midwave and longwave infrared single photon detectors is crucial for their emerging applications in spectroscopy, remote sensing, exoplanet detection, and free space quantum communications. However, existing sensors need to be operated at extremely low temperatures (0.08-0.9K) to reduce dark noise and hence require the use of advanced cryogenics such as dilution refrigerators or $^3$He cryogens, significantly limiting applications. Here we propose a vortex-engineering approach based on a hybrid phase transition in a ferromagnet/superconductor bilayer to increase the operating temperature of infrared single photon detectors up to 3.75K. We show that the introduction of a ferromagnetic layer produces a local magnetic field which impedes vortex crossing in the superconductor, reducing dark noise. When a single photon is incident, the photon-induced hotspot causes an order-to-disorder transition in the ferromagnet, leading to a vortex-induced phase transition in the superconducting layer. By engineering the ferromagnet's Curie temperature to be close to the device's operating temperature, single photon sensitivity can be achieved at increased operating temperatures. We predict at midwave/longwave infrared wavelengths (3-14$\mu$m) the operating temperature can be raised to 3.25-3.75K, enabling significantly simpler cooling systems.

阅读与讨论 → 访问原文 →

17.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.15152

Can Agents Read the Room? Benchmarking Visual Social Intelligence in Multimodal Simulation

作者:

Shijun Wan ↗Xuehai Wu ↗Jiwen Zhang ↗Siyuan Wang ↗Zhongyu Wei ↗

Social interaction depends on both language and visible social signals, such as facial expressions, posture, gaze, and emotional shifts. Yet existing social-agent benchmarks are largely text-based and rarely test whether multimodal agents can use visual cues to guide interaction. We introduce \textsc{\benchmarkname{}}, a benchmark evaluating visual social intelligence in multimodal social simulation. It contains 240 scenarios, 585 role instances, and 2,340 role-task instances, combining aligned textual-visual evidence, structured role profiles, and four role-level tasks: expression task, characteristic task, interaction regulation task, and interaction outcome task. Evaluating seven recent MLLMs under verbalized-vision and direct-vision reveals a clear gap between local role enactment and interaction management: role-specific expression and conflict handling are near saturation, whereas interaction regulation and visually grounded outcome achievement remain substantially more difficult. The code is released at https://github.com/JunsWan/AgentViSS, and the dataset is available at https://huggingface.co/datasets/JunsWan/AgentViSS.

阅读与讨论 → 访问原文 →

18.

medRxiv (Medicine) 2026-06-18 DOI: HASH:9434e01da852cfeba2c5a6891bfb5670

Hard to Halt: Automation Bias in Agent-Driven Sequencing Prior Authorization Workflows

作者:

Nie ↗Chung ↗Waxler ↗Lee ↗Weng ↗Lewis ↗Ahimaz ↗Wang ↗Liu ↗

Purpose: Prior authorization (PA) for exome or genome sequencing is a time-consuming process that impedes timely rare disease diagnosis. Large language model-based browser agents offer potential for automating these workflows, but their clinical reliability remain uncharacterized. Methods: We developed a sandbox compromising a simulated ES/GS PA submission payer portal and a synthetic EHR containing 836 patient records spanning compliant profiles and deficient profiles with different types of issues. Gemini 3 Pro, Gemini 3 Flash, and Claude Opus 4.5 were evaluated on task completion rate, form completion accuracy, and appropriate withholding for deficient profiles. Results: Larger models achieved much higher task completion rates (Gemini 3 Pro 95.45%, Claude Opus 4.5 93.67%) compared to Gemini 3 Flash (56.05%), but nearly universally failed to withhold submission for deficient profiles whereas Gemini 3 Flash ironically demonstrated superior withholding performance (17.33%). In a non-agentic setting, Gemini 3 Pro correctly identified 91% of the issues in deficient profiles, indicating that withholding failure is attributable to the browser interaction rather than the model's reasoning limitations. Conclusion: Current LLM-based browser agents exhibit a systematic bias towards form submission that poses risks in PA workflows. A modular, multi-agent architecture with human supervision is necessary for a safe clinical deployment.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.CL) 2026-06-18 DOI: arXiv:2606.18947

Decoupling Search from Reasoning: A Vendor-Agnostic Grounding Architecture for LLM Agents

作者:

Emmanuel Aboah Boateng ↗Kyle MacDonald ↗Amardeep Kumar ↗Siddharth Kodwani ↗Sudeep Das ↗

Production LLM agents increasingly depend on real-time search, yet native search grounding bundles retrieval policy, provider choice, evidence injection, cost, latency, and generation behavior behind a single model-provider boundary. This coupling makes grounding hard to inspect, tune, reuse, or port, and can trigger Search-Induced Verbosity that breaks strict output contracts. We present Decoupled Search Grounding (DSG), a vendor-agnostic boundary that moves grounding outside the reasoning model through an MCP-compatible gateway, exposing provider routing, source-aware context rendering, configured fallback, retrieval-depth control, and exact plus semantic caching as first-class controls. Across five frontier models on SimpleQA, FreshQA, and HotpotQA, native search leads on recency-sensitive FreshQA, but DSG exposes a stronger frontier when control matters: on SimpleQA it nearly matches native accuracy (86.1% vs. 87.7%) at 91% lower search cost, preserves concise answer contracts, and reaches a 99.4% warm-cache hit rate with 68% lower latency. Deployed as a shared production grounding layer for large-scale agentic workloads with interchangeable models, DSG matches or slightly exceeds native-search accuracy on an e-commerce query-understanding (QIU) workload while cutting search cost by over 98%. Real-time grounding is best treated as an optimizable interface boundary, not a fixed model feature.

阅读与讨论 → 访问原文 →

20.

arXiv (math.PR) 2026-06-18 DOI: arXiv:2501.18466

A random recursive tree model with doubling events

作者:

Jakob E. Bj\"ornberg ↗C\'ecile Mailler ↗

arXiv:2501.18466v3 Announce Type: replace Abstract: We introduce a new model of random tree that grows like a random recursive tree, except at some exceptional "doubling events" when the tree is replaced by two copies of itself attached to a new root. We prove asymptotic results for the size of this tree at large times, its degree distribution, and its height profile. We also prove a lower bound for its height. Because of the doubling events that affect the tree globally, the proofs are all much more intricate than in the case of the random recursive tree in which the growing operation is always local.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.LG) 2026-06-12 DOI: arXiv:2606.12966

Circuit Synchronization Precedes Generalization: Causal Evidence from Fourier Structure in Grokking Transformers

作者:

Achyuthan Sivasankar ↗

arXiv:2606.12966v1 Announce Type: new Abstract: Grokking – where a transformer on modular arithmetic suddenly transitions from near-chance to near-perfect validation accuracy – is attributed to a Fourier circuit, but its timing, causal structure, and controllability remain poorly understood. We introduce the Frequency Synchronization Degree (FSD), a normalised, permutation-tested metric for Fourier circuit synchronisation requiring no prior circuit knowledge. Across nine modular addition configurations (primes p in {53, 71, 97, 113, 131}, three seeds), FSD synchronises 500-3,000 steps before grokking (mean lead +1,722 steps; all nine positive, sign-test p~0.004), and precedes a restricted-logit loss baseline (Nanda et al.'s excluded loss) in all nine cases, making it the earliest available predictor. We provide direct causal evidence that the inter-phase gap is a regularisation phenomenon: forking training at the FSD-ceiling step and varying weight decay lambda produces strictly monotone earlier grokking, with Delta_t proportional to 1/lambda. This law replicates across three primes (p in {53,97,131}; R^2=1.00 and R^2=0.99 for two clean cases), captured as Delta_t ~ C/lambda, consistent with (1/lambda)*log(||W_mem||/tau). Architecture ablations show an attention-only model groks with a strong FSD precursor; an MLP-only model never groks; a single-layer model's FSD lags, confirming the precursor is a multi-block circuit property.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.CV) 2026-06-11 DOI: arXiv:2606.11682

Parameter-Efficient Adapter Tuning for Tabular-Image Multimodal Learning

作者:

Jiaqi Luo ↗

Tabular-image multimodal learning aims to improve predictive modeling by jointly using structured tabular attributes and visual data. Although pretrained encoders provide strong modality-specific representations, full fine-tuning can be computationally expensive, while keeping encoders frozen may limit task-specific adaptation. We propose the Tabular-Image Adapter (TI-Adapter), a modality-specific adapter-based fine-tuning framework for efficient multimodal adaptation. TI-Adapter freezes the pretrained tabular encoder and learns an adapter after the extracted tabular embedding, while adapting the image branch with embedding-level and bottleneck-level adapters instead of full fine-tuning. Experiments on 20 tabular-image datasets show that TI-Adapter achieves competitive or better predictive performance than full fine-tuning while using substantially fewer trainable parameters. Ablation studies further demonstrate the importance of adapter placement for balancing performance and practical efficiency.

阅读与讨论 → 访问原文 →

23.

arXiv (CS.AI) 2026-06-17 DOI: arXiv:2606.17541

Offline Preference-Based Trajectory Evaluation

作者:

Fernando Diaz ↗

arXiv:2606.17541v1 Announce Type: cross Abstract: Offline evaluation of agentic systems often collapses trajectories to terminal success, discarding information about partial progress and inducing widespread ties, creating substantial statistical inefficiency by reducing effective sample size and weakening the ability to distinguish systems. We propose preference-based trajectory evaluation, which compares trajectories directly through temporal preferences over progress and time-to-return profiles. We find that, across diverse agentic and interactive benchmarks, standard success-based metrics produce tied comparisons on roughly 75% of instances, whereas trajectory-aware preferences reduce ties to roughly 35%, improving discriminative power, ranking stability, and data efficiency. Our results suggest that benchmark saturation, often attributed to poor data collection or problem difficulty, may also be explained by the choice of evaluation measure.

阅读与讨论 → 访问原文 →

24.

arXiv (CS.LG) 2026-06-19 DOI: arXiv:2604.06001

A deep learning framework for jointly solving transient Fokker-Planck equations with arbitrary parameters and initial distributions

作者:

Xiaolong Wang ↗Jing Feng ↗Qi Liu ↗Chengli Tan ↗Yuanyuan Liu ↗Yong Xu ↗

arXiv:2604.06001v2 Announce Type: replace-cross Abstract: Efficiently solving the Fokker-Planck equation (FPE) is central to analyzing complex parameterized stochastic systems. However, current numerical methods lack parallel computation capabilities across varying conditions, severely limiting comprehensive parameter exploration and transient analysis. This paper introduces a deep learning-based pseudo-analytical probability solution (PAPS) that, via a single training process, simultaneously resolves transient FPE solutions for arbitrary multi-modal initial distributions, system parameters, and time points. The core idea is to unify initial, transient, and stationary distributions via Gaussian mixture distributions (GMDs) and develop a constraint-preserving autoencoder that bijectively maps constrained GMD parameters to unconstrained, low-dimensional latent representations. In this representation space, the panoramic transient dynamics across varying initial conditions and system parameters can be modeled by a single evolution network. Extensive experiments on paradigmatic systems demonstrate that the proposed PAPS maintains high accuracy while achieving inference speeds four orders of magnitude faster than GPU-accelerated Monte Carlo simulations. This efficiency leap enables previously intractable real-time parameter sweeps and systematic investigations of stochastic bifurcations. By decoupling representation learning from physics-informed transient dynamics, our work establishes a scalable paradigm for probabilistic modeling of multi-dimensional, parameterized stochastic systems.

阅读与讨论 → 访问原文 →

25.

arXiv (CS.CL) 2026-06-11 DOI: arXiv:2606.11762

Automated Creativity Evaluation of Language Models Across Open-Ended Tasks

作者:

Min Sen Tan ↗Zachary Kit Chun Choy ↗Syed Ali Redha Alsagoff ↗Nadya Yuki Wangsajaya ↗Mohor Banerjee ↗Swaagat Bikash Saikia ↗Alvin Chan ↗

Large language models (LLMs) have achieved remarkable progress in language understanding, reasoning, and generation, sparking growing interest in their creative potential. Realizing this potential requires systematic and scalable methods for evaluating creativity across diverse tasks. However, most existing creativity metrics are tightly coupled to specific tasks, embedding domain assumptions into the evaluation process, and limiting scalability and generality. To address this gap, we introduce an automated, domain-agnostic framework for quantifying LLM creativity across open-ended tasks. Our approach separates the measurement apparatus from the creative task itself, enabling scalable, task-agnostic assessment. Divergent creativity is measured using semantic entropy, a reference-free and robust metric for novelty and diversity, validated against human annotations, LLM-based novelty judgments and baseline diversity measures. Convergent creativity is assessed via a novel retrieval-based multi-agent judge framework that delivers context-sensitive evaluation of task fulfilment with over 60% improved efficiency. We validate our framework in three qualitatively distinct domains: problem-solving (MacGyver), research ideation (HypoGen), and creative writing (BookMIA), using a broad suite of LLMs. Empirical results show that our framework reliably captures key facets of creativity, including novelty, diversity, and task fulfilment, and reveal how model properties, such as size, temperature, recency, and reasoning, impact creative performance. Our work establishes a reproducible and generalizable standard for automated LLM creativity evaluation, paving the way for scalable benchmarking and accelerating progress in creative AI.

阅读与讨论 → 访问原文 →

探索全球前沿学术脉络