Academic Intelligence · Curated Daily

探索全球前沿学术脉络

AcademicHub 汇聚顶级期刊与预印本平台的实时文献。定制您的专属科研雷达,利用大语言模型自动生成交叉领域文献分析简报。

01.
arXiv (quant-ph) 2026-06-16

Systematic Construction of Time-Dependent Hamiltonians for Microwave-Driven Josephson Circuits

arXiv:2512.20743v4 Announce Type: replace Abstract: Time-dependent electromagnetic drives are fundamental for controlling complex quantum systems, including superconducting Josephson circuits. In these devices, accurate time-dependent Hamiltonian models are imperative for predicting their dynamics and designing high-fidelity quantum operations. Existing numerical methods, such as black-box quantization (BBQ) and energy-participation ratio (EPR), excel at modeling the static Hamiltonians of Josephson circuits. However, these techniques do not fully capture the behavior of driven circuits stimulated by external microwave drives, nor do they include a generalized approach to account for the inevitable noise and dissipation that enter through microwave ports. Here, we introduce numerical techniques that leverage classical microwave simulations, efficiently executable in finite-element solvers, to obtain the time-dependent Hamiltonian of microwave-driven superconducting circuits with arbitrary geometries under charge, flux, or mixed electromagnetic modulation. Importantly, our techniques do not rely on a lumped-element description of the superconducting circuit, in contrast to previous approaches to tackling this problem. We demonstrate the versatility of our approach by characterizing the driven properties of realistic circuit devices in complex electromagnetic environments, including coherent dynamics due to charge and flux modulation, as well as drive-induced relaxation and dephasing. Our techniques offer a powerful toolbox for optimizing circuit designs and advancing practical applications in superconducting quantum computing.

02.
arXiv (CS.AI) 2026-06-19

Bring My Cup! Personalizing Vision-Language-Action Models with Visual Attentive Prompting

arXiv:2512.20014v3 Announce Type: replace-cross Abstract: While Vision-Language-Action (VLA) models generalize well to generic instructions, they struggle with personalized commands such as "bring my cup," where the robot must act on one specific instance among visually similar objects. We study this setting of manipulating personal objects, in which a VLA must identify and control a user-specific object unseen during training using only a few reference images. To address this challenge, we propose Visual Attentive Prompting (VAP), a simple-yet-effective training-free perceptual adapter that equips frozen VLAs with top-down selective attention. VAP treats the reference images as a non-parametric visual memory, grounds the personal object in the scene through open-vocabulary detection and embedding-based matching, and then injects this grounding as a visual prompt by highlighting the object and rewriting the instruction. We construct two simulation benchmarks, Personalized-SIMPLER and Personalized-VLABench, and a real-world tabletop benchmark to evaluate personalized manipulation across multiple robots and tasks. Experiments show that VAP consistently outperforms generic policies and token-learning baselines in both success rate and correct-object manipulation, helping to bridge the gap between semantic understanding and instance-level control.

03.
medRxiv (Medicine) 2026-06-17

MedAgent: A Retrieval-Augmented Clinical Decision Support Agent with Verifiable Evidence Grounding for Evidence-Based Medicine

Evidence-based medicine demands clinical answers that are not only fluent and medically plausible, but also anchored in traceable evidence, tailored to patient-specific clinical questions, sensitive to the hierarchy of evidence, and respectful of clinical safety boundaries. While general-purpose large language models (LLMs) exhibit strong medical language generation ability, they tend to lean on parametric memory, underuse retrieved evidence, hallucinate citations, conflate evidence levels, and draw conclusions that are not fully supported by the underlying literature. Such limitations pose particular risks in clinical decision support, where answer reliability, evidence traceability, and reasoning consistency are paramount. To address these issues, we present MedAgent, an evidence-based medical agent trained through an end-to-end pipeline that integrates supervised fine-tuning (SFT) cold start, reward modeling, and Group Relative Policy Optimization (GRPO). The agent is designed to execute a structured workflow encompassing clinical question understanding, PICO extraction, evidence retrieval, evidence stratification, citation-grounded answer generation, and quality evaluation. Specifically, a Qwen2.5-14B-Instruct backbone is first cold-started on 200 human-verified agent trajectories, equipping it with tool invocation, PICO parsing, structured response generation, and citation faithfulness. Next, a Qwen2.5-7B reward model is trained on 2{,}099 pairwise preference samples to provide semantic-level quality signals for evidence-based responses. Finally, GRPO reinforcement learning is conducted in a retrieval-augmented agent environment, where every rollout involves real evidence retrieval and is scored jointly by rule-based rewards and reward-model signals. To avoid over-reliance on training rewards, we further construct an independent evidence-based medical evaluation benchmark, MedTrustBench, which contains 200 clinical questions spanning 10 specialties and four difficulty levels. Each question is annotated with standardized PICO elements and rubric-based scoring criteria. The benchmark includes 1{,}187 rubrics across seven dimensions: question relevance, evidence hierarchy, evidence quality and timeliness, evidence-answer consistency, completeness and depth, logical rigor, and medical terminology. Under an identical RAG pipeline, retrieval tool, retrieval configuration, and evaluation protocol, MedAgentv17 attains 78.6 points, outperforming GPT-4.1 (75.3) and approaching GPT-5.4 (80.3). These results show that a 14B domain-aligned model can surpass strong general-purpose baselines on specialized evidence-based medical reasoning, while delivering practical advantages in cost, privacy, controllability, and hospital-oriented private deployment. The model and associated datasets are publicly released at https://www.modelscope.cn/profile/InfoxmedModel

04.
arXiv (CS.LG) 2026-06-11

Urban Heat MiniCubes: An AI-Ready dataset for urban heat research

arXiv:2606.11534v1 Announce Type: cross Abstract: Urban heat is amplified by impermeable surfaces and heterogeneous built environments, yet street-level variability remains difficult to quantify because multi-sensor observations are rarely available in consistent, analysis-ready form at the necessary spatiotemporal scales. We present "Urban Heat MiniCubes," a publicly available, FAIR-oriented dataset designed for machine learning applications in urban heat research. The dataset provides harmonized 90 x 90 km gridded data cubes for 48 cities in the Western Hemisphere spanning 2022-2023, with variables reprojected and collocated to a common grid to reduce preprocessing (e.g., reprojection, resampling, and spatiotemporal alignment). Urban Heat MiniCubes includes two complementary modalities: (i) higher-spatial-resolution, lower-frequency observations from Landsat 8/9 (e.g., surface reflectances) and Sentinel-1 (e.g., synthetic aperture radar backscatter), and (ii) higher-temporal-frequency, coarser observations from GOES-R (e.g., longwave infrared brightness temperatures) and a microwave land surface temperature product. We document variables and metadata and provide technical assessment using inter-variable analyses and autoencoder-based reconstruction-error summaries across pixel classes (e.g., water and cloud). Potential use cases and limitations are also discussed.

05.
medRxiv (Medicine) 2026-06-18

Looked but didn't see: inattentional blindness and yes-bias confabulation in vision-language models

Previous work showed that many participants fail to notice a gorilla in a video of people playing basketball. Another study found that 83% of trained radiologists failed to report a gorilla figure inserted into a chest CT nodule-search task, even though eye-tracking revealed that most observers had foveated the figure. We ask whether a similar phenomenon exists in contemporary vision-language models (VLMs). We find that (i) VLMs are capable of spotting the gorilla in both still-frame images and videos of lung CT scans; (ii) models display inattentional blindness, which varies according to model generation and type of stimulus presented; (iii) Gemini-3.1-Pro outperforms most other flagship and open-weight VLMs at identifying the presence or absence of the gorilla. We additionally ran a segmentation experiment utilizing two different model classes: a generalist (SAM 3), which found the gorilla but produced little to no results for anatomy-based prompts; a medical specialist (BiomedParse), which produced more promising anatomy-based results but flagged "gorilla" on gorilla-free control videos on 82% of frames. The behavioral signature of inattentional blindness reproduces in VLMs, but a unique confabulation failure mode means that any "did the model see X" claim requires signal-detection analysis with a matched-control false-alarm baseline.

06.
arXiv (CS.AI) 2026-06-12

Who Pays the Price? Stakeholder-Centric Prompt Injection Benchmarking for Real-world Web Agents

arXiv:2606.13385v1 Announce Type: cross Abstract: Web agents driven by large language models (LLMs) are increasingly deployed in real-world environments, where they operate over untrusted web content and execute actions with direct consequences. This makes them vulnerable to prompt-injection attacks, in which seemingly benign content embeds adversarial instructions that manipulate agent behaviour. Existing security benchmarks adopt an attack-centric perspective, focusing on the technical feasibility of injections while overlooking the nuanced distribution of resulting harms. In practice, however, prompt-injection risk is victim-dependent: a single exploit can produce asymmetric consequences for different stakeholders, and the same attack pattern may exhibit substantially different effectiveness depending on whom it targets. To capture these properties, we introduce \sysname, a stakeholder-centric benchmark to systematically categorize and attribute harm in real-world web agent systems. It distinguishes between affected entities (e.g., user, seller, platform), decomposes the attacks into concrete objectives, and evaluates each case with complementary outcome- and process-level metrics. Our results reveal substantial and heterogeneous vulnerabilities: not a single attack objective is reliably resisted by current agents, and failures distribute across qualitatively distinct modes ranging from stealthy parasitism (attack succeeds without disrupting the user's delegated task) to misaligned disruption (task disrupted without attack success) and compounded failure (both adversarial objective and task integrity simultaneously violated). These patterns are missed by conventional evaluation, highlighting the need for stakeholder-aware assessment of LLM-based agents in real-world deployments. Benchmark is available at https://github.com/StakeBench/SBC.

07.
arXiv (CS.AI) 2026-06-12

ToolSense: A Diagnostic Framework for Auditing Parametric Tool Knowledge in LLMs

arXiv:2606.12451v1 Announce Type: new Abstract: Large language models deployed as agents over large tool catalogs face a critical tool-retrieval bottleneck. As embedding-based retrieval approaches rely on compact encoders that may under-capture specialized tool semantics, parametric tool retrieval addresses this by encoding each tool as a virtual token appended to the LLM vocabulary, fine-tuned in two stages (memorization then retrieval SFT) to use the LLM as a retriever, achieving strong performance on standard ToolBench retrieval benchmarks. Yet these benchmarks use verbose, fully-specified queries, and their evaluation applies constrained decoding that restricts outputs to valid token paths, neither reveals whether the model actually understands its tools. We introduce ToolSense, an open-source LLM-powered diagnostic framework that takes any tool catalog as input and automatically generates three benchmarks: a Realistic Retrieval Benchmark (RRB) with queries at three ambiguity tiers, an MCQ probing benchmark, and a QA probing benchmark. Applying ToolSense to ToolBench (~47k tools) and evaluating five parametric model training configurations reveals a knowledge-retrieval dissociation: on RRB queries, several configurations collapse by ~50-64 percentage points compared to fully-specified ToolBench benchmarks, falling below the embedding-model baseline. Additionally, despite strong retrieval performance, some models score near-random on factual probes, suggesting a knowledge-retrieval dissociation. We open-source the ToolSense framework and the ToolBench diagnostic benchmarks at https://github.com/SAP/toolsense.

08.
arXiv (CS.CV) 2026-06-17

ERQA-Plus: A Diagnostic Benchmark for Reasoning in Embodied AI

Generalist embodied agents require more than object recognition: they must reason about spatial relations, actions, procedures, human intentions, environmental constraints, and commonsense consequences from situated visual observations. Yet existing visual and embodied question answering benchmarks often provide limited control over the reasoning dependencies being tested, making it difficult to distinguish grounded embodied reasoning from shortcut-driven visual or linguistic pattern matching. We present ERQA-Plus, a diagnostic benchmark for reasoning in embodied AI. ERQA-Plus contains 1,766 question-answer instances grounded in 711 robot-centric images and organized according to a structured taxonomy spanning perceptual, action-centric, social-interaction, navigation-environmental, and contextual commonsense reasoning. The dataset is constructed using a multi-stage generation and validation pipeline that combines taxonomy-guided question generation, automatic quality judging, iterative revision, and human assessment to improve visual grounding, answer validity, and reasoning quality. We benchmark representative general-purpose vision-language models and embodied models, including LLaVA-NeXT-8B, Prismatic-7B, MiniCPM-V-4.5-8B, Qwen3-VL, RoboRefer-8B, and RoboBrain2.5-8B. Although the strongest model, Qwen3-VL-32B, achieves 83.4% overall accuracy and 61.4 SBERT score, category-level results reveal persistent weaknesses in spatial reasoning, procedural reasoning, event prediction, and intention inference. ERQA-Plus therefore provides a fine-grained evaluation framework for measuring not only whether embodied agents answer correctly, but also which forms of embodied reasoning they can and cannot perform reliably. The dataset is available https://huggingface.co/datasets/huggingdas/erqa-plus and the project page at https://github.com/LUNAProject22/erqa-plus.

09.
arXiv (CS.CV) 2026-06-15

RepFusion: Leveraging Multimodal Priors for Denoising in Representation Space

Large language models (LLMs) are widely used in text-to-image (T2I) systems, but they are typically limited to text encoding, while denoising is handled by newly trained generative backbones. The emergence of representation autoencoders (RAEs) shifts the generation target toward semantically structured visual representations, creating a latent space that is more compatible with pretrained LLM priors. Inspired by multimodal LLMs (MLLMs), where an MLP projector is sufficient to align clean visual representations with a pretrained LLM, we repurpose the MLLM itself as a noisy representation encoder, extending this mechanism from clean to noisy inputs. We present RepFusion, which uses the resulting MLLM outputs as the conditioning signal for a diffusion transformer. In controlled comparisons at similar inference budgets, RepFusion outperforms baselines that devote comparable capacity to newly initialized denoisers. These results demonstrate that MLLMs provide strong priors for denoising visual representations and that, by conditioning on evolving noisy representations, test-time compute can be productively spent on repeated MLLM conditioning in modern T2I systems.

10.
arXiv (CS.CL) 2026-06-16

KVEraser: Learning to Steer KV Cache for Efficient Localized Context Erasing

Post-hoc context erasing over the KV cache is challenging because a local edit has a global consequence: once a span has been processed, its influence propagates into the cached states of all subsequent tokens. This issue arises naturally in long-context LLM applications, where stale retrieved facts, incorrect tool observations, retracted user preferences, or harmful prompt injections may be identified only after prefill. Exact erasing must then recompute all tokens after the deleted span, making its computational cost depend on suffix length rather than erased-span length. We introduce KVEraser, a learned KV-cache editing method for efficient localized context erasing. Given a processed context and a span to remove, KVEraser replaces only the KV states of the erased interval with learned steering states while reusing the remaining cache unchanged. To learn a transferable erasing mechanism, we build a two-stage training pipeline: generic span-neighbor pre-training teaches the eraser to suppress the influence of the erased span, while task-specific fine-tuning adapts this capability to downstream scenarios. Experiments show that KVEraser nearly matches full recomputation in post-erasure performance on in-domain tasks across 1K–32K context lengths, while its latency increases by only 24% compared with a 17.6x increase for full recomputation. KVEraser also generalizes to unseen long-document QA tasks with harmful factual distractors, achieving the best performance among approximate baselines with a 3–4x speedup over full recomputation.

11.
arXiv (CS.CL) 2026-06-11

MemToolAgent: Leveraging Memory for Tool Using Agents Based on Environment and User Feedback

Modern large language model (LLM) agents can use external tools to help users solve complex tasks. However, for problems that require learning from long-term historical events or from previous agent-environment interactions, LLM agents are required to use memory mechanisms to store and retrieve experiences. While sophisticated memory systems exist for dialogue agents, few studies have empirically examined how to improve agents' tool-using capabilities through past user-agent conversations. We propose MemToolAgent, a framework that improves tool use through memory management. Our approach contains a memory extraction module that processes past experiences into structured memory entries, and a retrieval module that dynamically selects a subset of the stored memory entries. This enables more personalized and accurate responses aligned with user preferences and feedback without requiring LLM fine-tuning. In summary, this work has three main contributions: (1) a unified memory entry format that improves both general-purpose and personalized tool use without LLM fine-tuning, (2) a reflection-based memory extraction that uses environment and user feedback to distill wrong executions into critiques to store, and (3) a retrieval module that chooses how many past experiences to use based on the memory similarity distribution. MemToolAgent achieves 29%, 80%, and 17% relative improvements compared to strong baselines on the WorkBench, NESTFUL, and PEToolBench benchmarks, respectively.

12.
arXiv (CS.CL) 2026-06-11

When Roleplaying, Do Models Believe What They Say?

Language models can state that "the Earth orbits the Sun" and, when role-playing Aristotle, assert the opposite. Recent work argues that persona adoption is fundamental to how language models operate, with models constantly selecting the most appropriate persona for a given context. Does such role-playing merely change the model's outputs, or does it also affect what the model internally represents as truthful? We study this question with linear truth probes, applying them to LLMs role-playing historical personas whose likely beliefs differ from modern consensus. For each persona, we compare false claims the persona would likely have endorsed (*era-believed*) with topic-matched false claims they would not have endorsed (*era-false*). Across prompting, in-context learning, and supervised fine-tuning, persona induction suppresses era-believed statements less than equally false alternatives, yet they remain classified as false overall. Role-play therefore shifts what these models say more than what they internally represent as true. We contrast this with models trained on harmful advice that exhibit Emergent Misalignment (EM). Across three model families (Qwen 2.5 14B, Qwen 3 8B, and Llama 3.3 70B), their false claims move substantially toward the true region of probe space, are defended under challenge roughly half the time versus about a sixth for role-play, and are used in downstream reasoning. Role-play and Emergent Misalignment thus are points on a spectrum of belief internalization, where role-play changes what a model says with little representational change, while Emergent Misalignment shifts the internal representation of false claims without fully marking them as true.

13.
arXiv (CS.AI) 2026-06-11

Embodied-BenchClaw: An Autonomous Multi-Agent System for Embodied Spatial Intelligence Benchmark Construction

arXiv:2606.11909v1 Announce Type: new Abstract: Benchmarks are essential for evaluating embodied spatial intelligence, yet their construction is labor-intensive, hard to reuse, and difficult to maintain. Existing embodied benchmarks are often static and may quickly become saturated as models improve, limiting their ability to distinguish new capabilities. We propose Embodied-BenchClaw, an autonomous agentic system for constructing embodied spatial intelligence benchmarks. Given a user-specified evaluation intent, Embodied-BenchClaw automatically produces a complete and continually updatable benchmark package through a five-stage pipeline: intent blueprinting, data collection, structuring and cleaning, benchmark synthesis, and evaluation reporting. The pipeline is coordinated by three agents for planning, construction, and evaluation. To improve reusability and reliability, Embodied-BenchClaw introduces an extensible Skill Library and process quality control, enabling benchmark construction to be composable, verifiable, and repairable. We instantiate multiple benchmarks covering indoor spatial reasoning, outdoor spatial reasoning, robotic manipulation, quadruped robot navigation, UAV/aerial-view understanding, and static benchmark enhancement. These benchmarks span diverse embodied carriers, data sources, and spatial capabilities. Experiments with human evaluation, judge-based assessment, consistency checks, cost analysis, and ablations show that Embodied-BenchClaw can construct verifiable, executable, maintainable, and diagnostically useful embodied spatial benchmarks with reduced manual effort.

14.
arXiv (CS.CV) 2026-06-16

Exact Posterior Score Estimation for Solving Linear Inverse Problems

Diffusion and flow-based models learn powerful data priors by training a denoiser to reverse Gaussian corruption. To use this prior to solve a linear inverse problem, one needs to sample from the posterior, but the score that the prior provides is the unconditional score, not the posterior score. Existing methods either steer a fixed pretrained denoiser with approximate measurement-matching corrections, or train a conditional restoration model that abandons the denoising structure of the prior. We derive the exact posterior score in closed form for linear Gaussian inverse problems under general Gaussian interpolants, and show that posterior sampling reduces to a denoising problem at an operator-dependent shifted pivot under an anisotropic noise covariance. We turn this identity into Exact Posterior Score (EPS), a denoising training objective that preserves the input/output structure of standard pretraining and can therefore be trained from scratch or fine-tuned from a pretrained denoiser. At inference, EPS uses the same sampler as the underlying backbone, with no likelihood gradients or projections. We evaluate EPS on five linear inverse problems across FFHQ and ImageNet, where it outperforms training-free and training-based baselines on fidelity, perceptual, and distributional metrics, while using roughly an order of magnitude fewer denoiser evaluations than gradient-based posterior samplers.

15.
medRxiv (Medicine) 2026-06-22

Three multimodal large language models fail at clinically actionable breast pathology in three different directions

Background. Breast cancer treatment depends on histopathological features, such as grade and receptor-defined subtype; however, specialist pathologist access is constrained when the workforce is limited. Commercial multimodal large language models (MLLMs) accept hematoxylin and eosin (H&E) image tiles through paid interfaces without local hardware or fine-tuning. However, prior pathology evaluations addressed only coarse tasks. Whether they reach treatment-determining accuracy and whether vendors agree remain unclear. Methods. We aimed to evaluate three vendor-designated flagship MLLMs (Claude Sonnet 4.6, Gemini 2.5 Pro, GPT-5.5) in 427 invasive breast cancer cases. Each case went to all three with identical H&E tiles and prompts, and the subtype was inferred in the second call. The reference was an institutional sign-out report of an immunohistochemistry-derived subtype. We calculated the concordance, sensitivity, specificity, Cohen's kappa, and pairwise McNemar and Bowker tests. Findings. Claude ranked highest by raw histologic-type concordance but lowest by kappa, classifying all 23 lobular and seven micropapillary carcinomas as invasive breast carcinoma of no special type. The models anchored the Nottingham grade to three modal grades. None of the models reliably identified human epidermal growth factor receptor 2-positive disease. The failure direction was vendor-specific: Claude and GPT-5.5 were under-detected, whereas Gemini was over-called. Twelve prompt variants (4,056 calls) did not recover sensitivity. Interpretation. No current commercial MLLM reaches deployment-ready accuracy for any treatment-determining feature of breast pathology. As each vendor fails in its own fixed direction, changing vendors alters the type of error rather than removing it; therefore, the value of these models is assistive rather than autonomous. At USD 0.20-0.50 per case, they may serve as supervised draft generators that leave the diagnosis with the pathologist.

16.
arXiv (CS.AI) 2026-06-24

Agentic AI for Bilevel Long-Term Optimization of Policy-Driven Physical Layer Systems

arXiv:2606.24416v1 Announce Type: new Abstract: Network operators' changing policies, service requirements, and stringent real-time constraints render existing methods designed with fixed objectives and constraints ineffective. This paper presents Agentic long-term performance optimization (Agentic-LTPO), a nested bilevel optimization framework that can be applied to adaptive physical layer problem configuration. The key idea is to employ agentic AI to generate upper-level configurations in a bilevel optimization structure, where evolving operator policies, environment summaries, and historical experiences are translated into structured lower-level optimization problem configurations. The lower level solves the problems with updated configurations for real-time physical-layer decisions. Considering cell-free MIMO beamforming as a use case, we embody Agentic-LTPO by designing a new multi-agent decision process with retrieval-augmented experience-based verification in the upper level, together with a closed-form beamformer in the lower level. Experiments demonstrate that Agentic-LTPO exhibits strong adaptability to dynamic operator policies and effectively enhances the system's long-term performance by 57.2% compared to traditional methods.

17.
medRxiv (Medicine) 2026-06-15

Beyond the Apnea-Hypopnea Index: Physiological and Demographic Predictors of Excessive Daytime Sleepiness in Obstructive Sleep Apnea

Excessive daytime sleepiness (EDS) is a common but inconsistently predicted symptom of obstructive sleep apnea (OSA). OSA is typically diagnosed with polysomnography (PSG), and the current standard for severity assessment is the apnea-hypopnea index (AHI). AHI has many limitations, including its inability to explain physiological mechanisms or reflect variability in patient symptoms, such as EDS. This retrospective study aims to find physiological and demographic parameters that better predict EDS in patients with OSA and to evaluate whether these parameters outperform AHI using PSG data from the Mount Sinai Integrative Sleep Center. Clinical variables used to predict EDS included arousal index (AI), average oxygen desaturation during sleep, average heart rate during sleep, and AHI, along with demographic variables including age, sex, and BMI. Hypothesis tests, logistic regression models, and decision tree classifier models were performed on the data to discriminate sleepy from nonsleepy patients as determined by an Epworth Sleepiness Scale (ESS) score [≥] 10. AI and oxygen desaturation were found to be the most predictive physiological variables, and sex and BMI were found to be the most predictive demographic variables. The final decision tree model with these four variables outperformed the AHI in predicting EDS. These findings suggest that daytime sleepiness in OSA can be better explained by measures of apnea burden, oxygenation impairment, and patient demographics than by AHI alone, although these remain only modestly predictive. Future studies should focus on investigating more comprehensive physiological markers, multi-night sleep data, and more objective assessments of sleepiness.

18.
arXiv (CS.CL) 2026-06-17

Dynamic Rollout Editing for Reducing Overthinking in RL-Trained Reasoning Models

Long-form chain-of-thought reasoning can improve LLM performance on complex tasks, but models often continue generating unnecessary reasoning after a correct answer has emerged. We refer to this behavior as overthinking. We study this phenomenon from the perspective of GRPO-style reinforcement learning (RL) post-training, framing it as a training-time credit-assignment problem rather than merely a decoding-time stopping problem. In rollouts sampled at the onset of GRPO training, we observe that successful trajectories can exhibit a slightly higher degree of overthinking than unsuccessful trajectories for the same prompts. This early imbalance provides a starting point for an undesirable feedback loop: because GRPO assigns sequence-level credit, it cannot distinguish the solution-reaching prefix from the unnecessary continuation that lengthens a successful trajectory. Both receive positive update signal, allowing the initial imbalance to grow into more severe overthinking during training. To address this issue, we introduce Dynamic Rollout Editing (DRE), a training-time intervention for successful trajectories that continue thinking after answer emergence. DRE preserves the accepted verified prefix, edits the remaining thinking, and prefers the edited trajectory within the same RL group, weakening the preference signal for unnecessary thinking without penalizing the reasoning needed to reach the answer. Experiments across diverse tasks show the effectiveness of DRE.

19.
arXiv (CS.CV) 2026-06-12

Emerging Flexible Designs for Geospatial Multimodal Foundation Models

Foundation models are rapidly transforming Earth observation by enabling scalable pretraining across diverse unlabeled geospatial modalities. However, their architectural diversity ranging from encoder-only to encoder-decoder and masked autoencoding paradigms makes it challenging to assess performance trade offs in a consistent manner. In this work, we present an apples-to-apples comparison of leading FM architectures designed for geospatial multimodal reasoning, with a particular focus on flexibility across varied spectral band configurations. We standardize pretraining using identical self supervised learning objectives and training datasets, and evaluate all models under consistent parameterization on the GEOBench benchmark across classification and segmentation tasks. Our results offer new insights into the design trade-offs between model flexibility, modality alignment, and downstream task performance. By highlighting architectural strengths and limitations under controlled conditions, this study provides practical guidance for building next generation geospatial foundation models capable of robust multimodal reasoning.

20.
Nature (Science) 2026-06-22

Isotopic evidence for a cold and distant origin of 3I/ATLAS

Interstellar objects provide the only directly observable samples of icy planetesimals formed around other stars, and can therefore provide insight into the diversity of physical and chemical conditions occurring during exoplanet formation1−3. Here we report isotopic measurements of the interstellar comet 3I/ATLAS, which reveal an elemental composition unlike any Solar System body. The water in 3I/ATLAS is enriched in deuterium, at a level of D/H = (0.98 ± 0.06)%, which is more than an order of magnitude higher than in known comets, while its range of 12C/13C ratios (141–191 for CO2 and 123–172 for CO) exceeds typical values found in the Solar System, as well as nearby interstellar clouds and protoplanetary disks. Such extreme isotopic signatures indicate formation at temperatures  ≲ 30 K in a relatively metal-poor environment. When interpreted with respect to models for Galactic chemical evolution, the carbon isotopic composition implies that 3I/ATLAS may have accreted as long ago as 12 billion years, following a period of intense, early star formation. 3I/ATLAS thus represents a preserved fragment of an ancient planetary system.

21.
arXiv (CS.LG) 2026-06-16

Beyond Accuracy: Measuring Bias Acknowledgment in Chain-of-Thought Reasoning for Responsible AI Evaluation

arXiv:2606.15127v1 Announce Type: new Abstract: Reasoning models are increasingly used in settings where the final answer is not the only object of review: educational tools may show students intermediate steps, decision-support systems may require human oversight, and audit workflows may inspect traces for misleading or biased input. In such settings, two responses can receive the same final-answer score while differing in whether the trace explicitly flags injected biasing content. Accuracy-only evaluation collapses these cases. We study this gap as a measurement blind spot for responsible evaluation and introduce a minimal trace-level diagnostic with two axes: susceptibility (whether the bias breaks a previously correct answer) and acknowledgment (whether the trace contains a rubric-defined surface reference to the injected content). Across thousands of biased GSM8K trials, GPT-4o and Claude Sonnet~4 have similar susceptibility rates ($1.3\%$ vs.\ $1.2\%$) but substantially different acknowledgment rates ($13.0\%$ vs.\ $75.0\%$) under the same rubric.

22.
arXiv (CS.AI) 2026-06-16

Mask-Proof: An LLM-based Automated Data Curation Pipeline on Mathematical Proofs

arXiv:2606.15258v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly capable of mathematical problem solving and can even assist with research-level proofs, yet we still lack a scalable and reproducible way to measure step-level reasoning in long proofs across diverse sources. This evaluation gap limits trustworthy AI assistance in proof-certified scientific progress. Existing evaluations often emphasize final answers or rely on costly expert grading, while end-to-end proof generation remains open-ended and hard to verify automatically. We introduce Mask-Proof, a pipeline that turns real proofs into automatically checkable masked-step tasks. It masks key formula steps, provides the necessary surrounding context, and evaluates model reconstructions with an LLM-based equivalence judge using repeated votes for stability. The resulting Mask-ProofBench contains 292 curated problems across diverse research areas. Experiments with 17 models show that reasoning-enhanced models outperform standard models by 12% to 27%. Our evaluator achieves 96.8% agreement with expert annotators, enabling faithful, reproducible, and comparable measurement of step-level mathematical reasoning. Benchmark, annotations, and code are available at https://github.com/weating/Mask-Proof.

23.
arXiv (CS.AI) 2026-06-11

Reliability-Calibrated Edge-IoT Early Fault Warning for Rotating Machinery with a Physics-Guided Tiny-Mamba Transformer

arXiv:2601.21293v3 Announce Type: replace-cross Abstract: Industrial Internet of Things (IIoT) systems increasingly rely on distributed vibration sensing to support predictive maintenance of rotating machinery. In practical deployments, however, raw signal upload is costly and alarm decisions must be made locally under limited computation, changing operating conditions, and strict nuisance-alarm budgets. This paper presents a reliability-calibrated edge-IoT early-warning framework, in which a compact Physics-Guided Tiny-Mamba Transformer (PG-TMT) acts as the representation module and an extreme value theory (EVT) layer converts streaming anomaly scores into event-level alarm episodes. PG-TMT combines a depthwise-separable convolutional stem, a Tiny-Mamba state-space branch, and a lightweight local Transformer to capture transient, long-horizon, and multichannel degradation cues under batch-size-one inference. To improve auditability, temporal attention is projected to the frequency domain and softly aligned with analytical bearing fault-order bands. EVT calibration, dual-threshold hysteresis, and trimmed-tail fitting provide controllable false-alarm intensity even when healthy calibration data are imperfect. Experiments on CWRU, Paderborn, XJTU-SY, and an industrial pilot demonstrate that the proposed framework improves PR-AUC, reduces detection delay under a controlled nuisance-alarm budget, and remains robust to structured interference, metadata uncertainty, compound fault mixtures, and domain transfer. With a sub-1 MB footprint and Jetson p99 latency below 7 ms, the framework supports calibrated and interpretable early warnings for IIoT predictive maintenance.

24.
arXiv (CS.AI) 2026-06-17

Distributed General-Purpose Agent Networks: Architecture, Key Mechanisms, and Prototypes

arXiv:2606.17368v1 Announce Type: new Abstract: Large language models have accelerated the transition from passive conversational assistants to autonomous agents that can understand goals, plan actions, invoke tools, and execute multi-step tasks. Yet the capability of a single agent remains constrained by its local data, tool permissions, runtime environment, and governance boundary. This paper studies distributed general-purpose agent networks: open peer-to-peer networks in which heterogeneous agents deployed on personal devices, edge nodes, or autonomous computing environments can discover one another, establish trust, negotiate cooperation rules, and execute open-ended tasks. We argue that such networks cannot be obtained by simply combining existing peer-to-peer overlays with conventional multi-agent systems. Unlike traditional P2P networks, agent networks must propagate semantic declarations about intentions, capabilities, states, and cooperation constraints. We therefore propose a layered architecture centered on a protocol adaptation layer that connects upper-level task semantics with lower-level network operations. Based on this architecture, the paper identifies three core mechanism problems: semantic announcement propagation for collaborator discovery, verifiable identity and multi-topic reputation for cooperation governance, and semantic-gradient mechanism design for open task execution. For each problem, we present a technical route, including bodyless gossip with sequential logs, BAID-based identity binding with MG-EigenTrust reputation, and a Stackelberg-style mechanism-generation loop driven by semantic attribution feedback. We further report prototype overhead results for BAID-style tiered verification and mechanism-level simulations of MG-EigenTrust under cross-topic disguise-collusion attacks. The resulting framework provides a system-level foundation for open, trustworthy, and scalable agent collaboration.

25.
arXiv (CS.CV) 2026-06-18

LandslideAgent with Multimodal LandslideBench: A Domain-Rule-Augmented Agent for Autonomous Landslide Identification and Analysis

Intelligent landslide hazard interpretation is critical for disaster prevention, yet current paradigms struggle to simultaneously extract visual features and high-level geoscientific semantics, while general-purpose vision-language models (VLMs) suffer from perceptual limitations and domain hallucinations in complex geological scenarios. To address these challenges, we propose an instruction-driven agentic framework comprising three components. First, LandslideBench, a multimodal fine-grained dataset with seven subtype labels, high-resolution imagery, pixel-level masks, and high-quality textual descriptions, is constructed via multi-VLM cross-validation and interactive annotation. Then, LandslideVLM, a landslide-oriented VLM, is fine-tuned via LoRA on LandslideBench to enhance geological semantic understanding. Finally, LandslideAgent, a domain rule-enhanced agent taking LandslideVLM as its cognitive backbone, employs a dual-rule controller incorporating structured report metadata constraints and cross-validation identification constraints to regulate automated tool invocation. Experiments demonstrate that LandslideBench provides effective baselines across five mainstream models on fine-grained classification and semantic segmentation. LandslideVLM achieves accuracy improvements of 10.96%, 32.87%, and 15.91% on landslide discrimination, fine-grained classification, and semantic description quality, respectively. LandslideAgent further enables autonomous multi-source spatial data inference, realizing full-process intelligence for landslide identification and analysis.