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 (CS.CL) 2026-06-16

Surpassing Scale by Efficiency: A Compact 135M Parameter Foundational LLM Natively Adapted for the Bangla Language

While the NLP landscape is dominated by multi-billion parameter architectures, their deployment in low-resource, non-Latin scripts remains computationally prohibitive for edge configurations, mobile systems, and decentralized local hardware. This paper presents bangla-smollm-135m, a highly compact 135-million parameter decoder-only foundational model engineered explicitly for high-efficiency language modeling in the Bangla script. By leveraging a deterministic intersect-and-append token merging strategy between TituLLMs and SmolLM2-135M, the model overcomes subword script fragmentation without destabilizing early pretrained parameter states. In zero-shot multi-task benchmark evaluations (PIQA_bn, OpenBookQA_bn, CommonsenseQA_bn, and Bangla_MMLU), bangla-smollm-135m matches or outperforms models twice its size (Gemma-3-270m) and achieves parity with models in the 1B parameter tier. The model is available at rnnandi/bangla-smollm-135m

02.
arXiv (CS.AI) 2026-06-16

Medical world models: representing medical states, modelling clinical dynamics and guiding intervention policies

arXiv:2606.16721v1 Announce Type: new Abstract: Medical diagnosis and treatment are dynamic processes in which patient states evolve over time and clinical interventions alter future outcomes. Although current medical AI can detect disease, estimate risk and generate reports, many systems still return static labels or scores, offering limited insight into how illness may progress or how alternative interventions may reshape its trajectory. Medical world models adapt the world-model idea from artificial intelligence to healthcare by learning internal simulators of patient-state dynamics. Their long-term goal is to help clinicians anticipate deterioration, compare treatment-conditioned futures and tailor care to individual patients. Yet relevant work remains scattered across foundation models, longitudinal modelling, disease simulation, treatment-effect estimation, reinforcement learning and digital twins. To bridge this gap, this review outlines a roadmap for advancing medical AI from isolated diagnosis and prediction toward medical world models that simulate disease evolution and support intervention decisions. This roadmap is organized around three coupled capabilities: patient-state construction, clinical dynamics modelling and intervention decision support. Across representative systems, the comparison highlights what each capability contributes and how partial components can be integrated into more mature perception–dynamics–planning systems. Finally, we identify the challenges involved in turning plausible rollouts into clinically useful simulators. Related literature is available at https://github.com/1999kevin/awesome_medical_world_models.

03.
medRxiv (Medicine) 2026-06-17

Frequency-dependent cognitive effects of Deep Brain Stimulation in Parkinson's Disease: A Systematic Review and Meta-Analysis

Background: Subthalamic nucleus deep brain stimulation (STN-DBS) improves levodopa-induced motor complications and cardinal motor symptoms of Parkinson's disease (PD), but stimulation frequency may differentially shape outcomes. This is evident for axial and gait symptoms, which may respond differently to lower-frequency stimulation. Whether frequency-dependent effects extend to cognition remains unclear. Objective: To investigate the cognitive effects of DBS at distinct frequencies in PD. Methods: We conducted a systematic review and meta-analysis (PROSPERO - CRD42024618253). PubMed, Web of Science, and EMBASE were searched for studies assessing cognitive outcomes under different stimulation frequencies. Eight cognitive domains were defined: verbal fluency, cognitive flexibility, executive control, working memory, attention, processing speed, episodic memory, and time processing. Multilevel random-effects meta-analyses were performed, with effect sizes expressed as Hedges' g. Results: Forty-three studies met the inclusion criteria, the majority (n = 31) involving STN-DBS. Twenty-one STN-DBS studies, including 355 patients, were included in the meta-analysis. Compared with HFS ([≥] 130 Hz), lower frequencies (4-80 Hz) were associated with better verbal fluency (g = 0.27) and cognitive flexibility (g = 0.38), with consistent effects across sensitivity and leave-one-out analyses. Accuracy-based executive control measures also favored lower-frequency stimulation. OFF-stimulation comparisons showed a concordant pattern. Evidence for other targets (PPN and NBM) was limited. Conclusions: Lower-frequency STN-DBS was associated with modest benefits in specific cognitive domains compared with HFS. These findings highlight the need for future research to determine how frequency interacts with stimulation location and symptom-specific networks to shape cognitive and cognitive-motor outcomes in PD.

04.
arXiv (CS.AI) 2026-06-19

Toward Calibrated Mixture-of-Experts Under Distribution Shift

arXiv:2606.20544v1 Announce Type: new Abstract: Calibration aligns a model's predictive uncertainty with the frequencies of its empirical outcomes and is important for understanding and trusting reported probabilities. Recent work shows that enforcing calibration at the level of individual predictors can improve ensemble accuracy and calibration, with mixture-of-experts (MoE) models showing strong empirical improvements in particular; however, the conditions under which calibration helps MoE are not well understood. In this work, we study how MoE models behave under distribution shift, focusing on how routing mechanisms interact with expert-level calibration. We show that expert calibration is sufficient to ensure calibration of the overall model under a broad class of distribution shifts in hard-routed models, but is insufficient for calibrating soft-routed models. To address this, we propose an adversarial reweighting that penalizes calibration errors of the routed aggregate under distribution shift, and we demonstrate that it improves the accuracy-calibration tradeoff both on average and on difficult subsets of the data, across model classes, prediction tasks, and distribution shifts.

05.
arXiv (CS.CV) 2026-06-16

Keep It in Mind: User Centric Continual Spatial Intelligence Reasoning in Egocentric Video Streams

We introduce UCS-Bench, a dataset spanning 170+ hours of egocentric visual observations with 8.1K+ timestamped questions for diagnosing User-Centric Continual Spatial intelligence in egocentric video streams. UCS-Bench targets a new problem that emphasizes dynamic spatial reasoning, long-term memory, and their alignment with users' real-time locations. We propose DirectMe, a framework that incrementally constructs and maintains a structured spatial memory from streaming egocentric observations. DirectMe enables robust tracking and recall of object locations, all relative to the user's movement over time. By tightly coupling visual perception with memory updates and spatial reasoning, our approach supports long-horizon queries that require recalling interactions, resolving viewpoint-induced ambiguities, and adapting to dynamic scenes. Our experiments show that DirectMe significantly improves the spatial reasoning of leading multimodal LLMs; it also surpasses many spatially aware and long-form streaming video models. We hope our benchmark and solution will advance spatial intelligence research for egocentric AI assistants. Data and code are available at https://github.com/cocowy1/UCS-Bench.

06.
medRxiv (Medicine) 2026-06-17

Characterizing the genetic basis of Cardio-Renal-Metabolic multimorbidity using multivariate genomic modelling

Cardio-renal-metabolic multimorbidity (CRMM) encompasses interrelated conditions affecting the heart, kidneys, and metabolic systems. Although the genetics of individual components are well studied, their shared architecture remains unclear. Here, we performed the largest multi-ancestry multivariate GWAS of CRMM across seven biobanks, including individuals of European (EUR; neff = 353,130), African (AFR; neff = 75,436), and East Asian (EAS; neff = 164,373) ancestry. We identified 287 lead loci in EUR, 30 in AFR, and 202 in EAS. Cross-ancestry analyses revealed ancestry-specific signals and 24 shared loci mapping to FTO and TCF7L2. Drug-repurposing highlighted candidates used for type 2 diabetes and hypertension. Mendelian randomization supported causal links with diverse diseases, while polygenic risk scores showed improved prediction across ancestries. Collectively, these findings advance understanding of CRMM genetics and inform precision medicine.

07.
arXiv (quant-ph) 2026-06-17

Independent Chiral Control in Theory-Space Models:A Rank-Preserving Framework and Its Application to Neutrino Mass Generation

arXiv:2409.09033v3 Announce Type: replace-cross Abstract: We develop a general framework of rank-preserving, element-wise matrix transformations for engineering fermion mass hierarchies in theory-space constructions. We prove that preservation of massless modes requires the transformation function to be separable, $g_f(i,j)=g^{(L)}_f(i)g^{(R)}_f(j)$, which in turn enables independent control of left- and right-chiral zero-mode profiles directly at the level of the theory-space mass matrix. This formalism unifies and extends the clockwork mechanism, permits controlled deformation of Kaluza–Klein spectra, and enhances hierarchy generation in GIM-like fine-cancellation scenarios. As a concrete application, we show that in theory-space models for neutrino masses, suitable transformations allow sub-eV light neutrinos to arise from TeV-scale new physics with only $\mathcal{O}(40)$ additional fermionic sites, while remaining consistent with charged-lepton flavor-violation bounds. In contrast, the corresponding untransformed models asymptote at the MeV scale and cannot access the phenomenologically required regime without extreme field multiplicities or hierarchical parameters.

08.
medRxiv (Medicine) 2026-06-17

Proteomics Uncovers Cryptic JPH2 Loss in Paediatric Dilated Cardiomyopathy

Despite recent advances in next-generation sequencing, genetic diagnostic rates for dilated cardiomyopathy (DCM) remain low. Among paediatric DCM, causes are often heritable, with a greater frequency of de novo, recessive and syndromic causes of disease. Novel diagnostic methods are therefore required to solve monogenic cases. To assess the value of proteomics as a diagnostic tool for paediatric DCM, we obtained left ventricle myocardial samples from paediatric patients undergoing heart transplantation at the Royal Children's Hospital, Melbourne. We performed genome sequencing and proteomics and leveraged this multi-omics dataset to uncover the molecular cause of disease in a gene elusive proband. The proband carried a heterozygous JPH2 frameshift variant identified on clinical exome sequencing. However, proteomic analysis showed a pronounced downregulation of JPH2, suggestive of biallelic loss-of-function. Closer inspection of the genomic data revealed a large inversion (~8.34 Mb) with a breakpoint falling within intron 5 of JPH2 that displaces the 3'UTR from the coding transcript. The two variants were confirmed to be in trans using long read DNA sequencing, consistent with a diagnosis of JPH2 autosomal recessive DCM. Finally, we applied RNA sequencing with total RNA library preparation to show that transcripts containing a 3'UTR were reduced to ~10% relative to controls. As a proof-of-principle, we present the first reported use of proteomics from explanted cardiac tissue to provide a genetic diagnosis. Our methodology has broad relevance to patients with genetically unsolved Mendelian diseases, who might undergo organ transplantation as part of clinical management.

09.
arXiv (CS.CV) 2026-06-19

Cinematic Compositing Using Character-Environment-Harmonized Video Generation Models

Cinematic compositing aims to integrate green-screen characters into novel environments while maintaining physical and photometric realism. Previous methods often fail to capture the complex bidirectional interactions between characters and their surroundings, which we characterize as Character-to-Environment (C2E) physical interaction and Environment-to-Character (E2C) lighting harmonization. To address this, we propose an end-to-end video diffusion framework that jointly models C2E and E2C interactions, specifically handling the challenges of interactive props. Our approach introduces a tri-mask-guided architecture with RGB-D joint denoising to ensure physically consistent interactions among the character, props, and environment. We further develop an efficient prior-driven data curation pipeline to construct high-quality relighting pairs without expensive rendering. Finally, a reference-conditioned mechanism enables controllable environment synthesis and precise prop replacement. Extensive experiments demonstrate that our framework significantly outperforms existing methods in cinematic-quality dynamic video compositing.

10.
arXiv (quant-ph) 2026-06-12

Explicit Quantum Circuit Simulation of Nonlinear 1-Dimensional Fluid with Carleman-linearized Boltzmann Method

arXiv:2606.12770v1 Announce Type: new Abstract: Quantum computation of fluid dynamics has attracted growing attention as a key application of fault-tolerant quantum computers anticipated in the coming decade, with lattice Boltzmann methods emerging as a particularly promising approach. Explicit and efficient elementary-gate-level circuit simulations, however, have so far been demonstrated only in the linear case. Here we include the leading nonlinearity through second-order Carleman linearization of the one-dimensional Boltzmann equation, and demonstrate, via explicit quantum-circuit simulation, the preparation of the final-time state using a Taylor-expansion-based ODE solver based on the quantum singular value transformation. With this construction, we analyze the gate and qubit complexities, which scale logarithmically with the grid size, the nonlinearity captured by the higher-order Carleman linearization, and the practical utility of higher-order expansions in the Taylor ODE solver. The construction provides a concrete baseline for computational cost reduction and further developments such as extensions to higher dimensions, complex geometries, and the extraction of physical quantities, towards industrially useful quantum CFD.

11.
arXiv (CS.CL) 2026-06-17

MoSE: Mixture of Slimmable Experts for Efficient and Adaptive Language Models

Mixture-of-Experts (MoE) models scale large language models efficiently by sparsely activating experts, but once an expert is selected, it is executed fully. Hence, the trade-off between accuracy and computation in an MoE model typically exhibits large discontinuities. We propose Mixture of Slimmable Experts (MoSE), an MoE architecture in which each expert has a nested, slimmable structure that can be executed at variable widths. This enables conditional computation not only over which experts are activated but also over how much of each expert is utilized. Consequently, a single pretrained MoSE model can support a more continuous spectrum of accuracy-compute trade-offs at inference time. We present a simple and stable training recipe for slimmable experts under sparse routing, combining multi-width training with standard MoE objectives. During inference, we explore strategies for runtime width determination, including a lightweight test-time training mechanism that learns how to map router confidence/probabilities to expert widths under a fixed budget. Experiments on GPT-style models, various routing regimes, zero-shot downstream reasoning benchmarks, and continual pre-training adaptation of DeepSeek model show that MoSE matches or improves standard MoE at full width and consistently shifts the compute-quality frontier toward lower inference FLOPs. The code can be found at: https://github.com/tnurbek/mose.

12.
arXiv (CS.CL) 2026-06-15

MineExplorer: Evaluating Open-World Exploration of MLLM Agents in Minecraft

Multimodal large language models (MLLMs) have shown strong capabilities in perception, reasoning, and action generation. However, their ability to sustain exploration in dynamic open worlds remains unclear. Existing embodied and game-based benchmarks often compress interaction into short-horizon tasks or entangle success with domain-specific game mechanics. In this paper, we introduce MineExplorer benchmark for evaluating open-world exploration capabilities of MLLM agents in Minecraft. We first filter atomic tasks whose solutions rely heavily on Minecraft-specific knowledge to better reflect general open-world reasoning. Then we organize the benchmark around a ReAct-style capability formulation and compose atomic tasks into implicit multi-hop tasks. To further construct reliable instances, MineExplorer uses a multi-agent synthesis workflow that jointly designs task graphs, sandbox scenes, and rule-based milestone evaluators. Human evaluation shows that the multi-agent synthesis workflow produces significantly more reliable instances than a single-agent baseline. Experiments with advanced MLLM agents show that open-world exploration remains challenging, as strong models can handle many single-hop tasks but degrade sharply when hidden prerequisites must be coordinated over longer trajectories. Further analysis finds that task difficulty tracks agent completion, and larger models or thinking modes do not consistently translate into better performance. Code and dataset are available at https://github.com/Jometeorie/MineExplorer.

13.
arXiv (math.PR) 2026-06-19

Power-law hypothesis and (un)fairness of PageRank on undirected multi-type PAMs

arXiv:2606.19583v1 Announce Type: new Abstract: The preferential attachment model (PAM) describes the sequential growth of a network based on the "rich-get-richer" principle. Several versions of it have become established for modeling, e.g., citation networks, capturing a power-law degree distribution. Directed versions of the preferential attachment model where the edges are directed from the new to the old vertices have been the subject of extensive research. They have been shown to exhibit remarkable properties such as heavier tails for the limiting graph-normalized PageRank than for the in-degrees. By contrast, for the undirected version, we recently showed that PageRank has similar tails as the degree. In the present paper, we discuss the PageRank asymptotics for a multi-type version of the undirected PAM (here vertices have different colors), complementing previous results of Antunes, Bhamidi, Banerjee and Pipiras on the asymptotics of PageRank on similar directed multi-type or colored PAMs. Our studies are motivated by the aim to go beyond the rigid rule of edge orientation in directed preferential attachment models. As the main result, for the case of a finite set of colors, we show that the power-law hypothesis for PageRank is fulfilled also for the colored undirected PAM, where, by contrast to the directed case, the power-law exponent is color-dependent for some choices of the initial color distribution and the attractiveness function. For the specific case of a two-type model, we discuss implications of our results on fairness in sampling underrepresented nodes from the network.

14.
medRxiv (Medicine) 2026-06-17

Perceptions of aging well among older adults with heart failure: insights from a qualitative study

Background: Heart failure (HF) is a prevalent and often debilitating cardiovascular condition among older adults, frequently accompanied by multimorbidity, functional limitations, and the need to age in place. Traditional models of successful aging emphasize disease absence and preserved function, yet most individuals with HF live with ongoing symptoms and chronic health challenges. How older adults with HF define aging well, particularly across different socioeconomic contexts, remains underexplored. Objectives: To explore how older adults with HF conceptualize aging well and to identify perceived facilitators and barriers across more and less resourced New York City neighborhoods. Methods: We conducted semi-structured interviews with 20 adults diagnosed with HF residing in Manhattan and Brooklyn neighborhoods classified by 2019 United States Census data. Interviews were guided by Rowe and Kahn's model. Transcripts were analyzed using an inductive-deductive thematic approach and interpreted in alignment with the Healthy People 2030 framework. Results: Participants had a mean age of 69 years; 50% identified as Black and 50% were women. Despite functional limitations, 65% reported aging well. Five themes emerged: maintaining physical function, maintaining cognitive function, sustaining social relationships, avoiding pain, and promoting overall well-being. Avoiding pain and promoting well-being extended beyond traditional models. Neighborhood context shaped priorities, with financial stability emphasized in more affluent areas and social cohesion prioritized in less affluent communities. Conclusions: Older adults with HF frequently perceive themselves as aging well despite chronic illness, reframing successful aging beyond disease avoidance. These findings support a patient-centered, place-informed model of aging well with implications for healthcare delivery and policy.

15.
arXiv (CS.LG) 2026-06-16

Graph Learning Should Move Beyond Restrictive Views of Spectral and Message-Passing GNNs

arXiv:2602.10031v2 Announce Type: replace Abstract: Graph neural networks (GNNs) are commonly divided into message-passing neural networks (MPNNs) and spectral GNNs, reflecting two largely separate research traditions in machine learning and signal processing. While MPNNs have a precise definition, there is no widely accepted criterion for what makes a mapping a spectral GNN. Most existing work restricts spectral GNNs to layered architectures based on linear spectral filters. Under this restriction, we show that spectral and spatial GNNs have largely equivalent expressive power. To promote progress in the field, we propose a precise definition of spectral GNNs based on eigenbasis symmetries, in contrast to the definition of MPNNs via neighborhood permutation symmetries. We further argue that the two perspectives offer complementary strengths. MPNNs provide a natural language for discrete structure and expressivity analysis through tools from logic and graph isomorphism, while the spectral perspective offers principled tools for understanding smoothing, bottlenecks, stability, and community structure. Overall, we argue that progress in graph learning will be accelerated by clarifying the similarities and differences between these perspectives and by moving toward a unified theoretical framework.

16.
arXiv (math.PR) 2026-06-12

Characterizing metric-space-valued processes: separating classes and weak invariance principles for measure-theoretic inference

arXiv:2606.13084v1 Announce Type: cross Abstract: This article investigates stochastic processes taking values in metric spaces that lack a topological vector space structure, a regime characterized by intricate interplay between topological, geometric, and temporal dependence structures. It is formally established that spaces admitting an isometric Hilbertian embedding constitute a strict subclass within the much broader class of metric spaces possessing the ball property. While traditional kernel methods are susceptible to geometric distortion when the underlying space cannot be isometrically embedded into a Hilbert space, we bypass such limitations by exploiting a fundamental structural property inherent to this broader class; namely, that Borel probability measures are uniquely determined by their values on balls. These separating classes provide the foundation for the subsequently introduced measure-theoretic inference methodology. We derive uniform convergence of a family of time-dependent random measures, alongside weak invariance principles for the corresponding nonstationary random fields. This framework explicitly exposes how dependence and geometric complexity influence sample path regularity. Furthermore, because the rapid decay of small-ball probabilities can prohibit the existence of limiting distributions for supremum-based discrepancy measures, we develop $L^p$-based alternatives. By directly leveraging the introduced convergence results, this approach circumvents the need for higher-order $U$-process formulations. Finally, for spaces that do admit an isometric Hilbertian embedding, and where $U$-processes naturally arise, we establish limit theory for both degenerate and nondegenerate multi-parameter $U$-processes, and demonstrate that local discrepancy tests maintain asymptotic stability under dynamic parameter regimes.

17.
medRxiv (Medicine) 2026-06-15

Semantic Embeddings and the Peripheral Transcriptome in Ischemic Stroke: Connecting Molecular Signatures to NANDA-I Diagnoses

Objective: To construct and evaluate, in an exploratory manner, a pathophysiologic rationale link- ing biological pathways derived from the peripheral transcriptome in ischemic stroke (IS) to nursing diagnoses in the NANDA-I 2024-2026 taxonomy, while emphasizing that this association is not di- rect, deterministic, or automatically inferable from textual similarity with large language models (LLMs). Methods: A computational study was conducted using public secondary data from the Gene Ex- pression Omnibus series GSE16561, which includes 63 peripheral blood samples: 39 from indi- viduals with IS and 24 from healthy controls. The pipeline integrated transcriptomic analysis and functional enrichment, semantic mapping through ClinicalBERT embeddings, and mechanistic and clinical-conceptual judgment using Claude Sonnet 4.6 as a judge. The judgment stage was treated as the central interpretive layer, designed to mediate the transcriptome, pathophysiology, functional manifestation, and NANDA-I diagnosis. Results: The analysis identified a bimodal transcriptomic pattern, with activation of pathways re- lated to innate immunity and suppression of pathways related to adaptive immunity. Semantic map- ping generated 158 pathway-diagnosis pairs. The Spearman correlation between cosine similarity and the mechanistic score was negative and statistically significant (rho = -0.243; p = 2.09e-03), but weak in magnitude. This effect size indicates that semantic similarity explained less than 6% of the variance in mechanistic plausibility, reinforcing the insufficiency of embeddings as a stand- alone criterion. Of the 158 pairs, 14 were classified as high concordance, 8 as moderate, and 136 as divergent. Conclusion: The main value of this study lies in demonstrating that translating biological pathways into nursing diagnoses requires pathophysiologic, functional, and clinical-conceptual mediation. The prioritized pairs represent mechanistically plausible hypotheses for future research, without implying causality, direct clinical confirmation, or immediate care recommendations.

18.
arXiv (CS.AI) 2026-06-18

MagpieTTS-LF: Inference-Time Long-Form Speech Generation Without Training on Long-Form data

arXiv:2606.18485v1 Announce Type: cross Abstract: Neural Text-to-Speech (TTS) systems achieve remarkable quality on short utterances but long-form speech generation shows prosodic drift, speaker inconsistencies and sentence boundary artifacts. Existing approaches either compress sequences, increase context length or naively concatenate independently synthesized chunks. We present an inference-time approach called MagpieTTS-LF that enables MagpieTTS to produce coherent long-form speech without model retraining. Our method introduces three key innovations: (1) soft attention priors to guide monotonic alignment while preserving past and future context; (2) a stateful inference algorithm that maintains context across sentence chunks, ensuring prosodic continuity; (3) history-aware text encoding that uses past text for discourse-level prosodic planning. Experiments on long texts show significant improvements in long-range intelligibility, prosodic coherence, speaker consistency, and boundary naturalness compared to other baselines.

19.
arXiv (CS.AI) 2026-06-18

Spotlight: Synergizing Seed Exploration and Spot GPUs for DiT RL Post-Training

arXiv:2606.19004v1 Announce Type: cross Abstract: Reinforcement learning (RL) post-training of Diffusion Transformers (DiTs) is prohibitively expensive, requiring thousands of high-end GPUs. Existing works explore two directions to reduce cost: seed exploration improves training convergence by selecting high-contrast samples, yet adds compute to the critical path; spot GPUs offer 69–77\% lower cost, yet sit idle during training because DiT rollouts finish nearly simultaneously, which prevents LLM-style pipelining of rollout with training. Spot preemptions further break Sequence Parallelism (SP) groups, fragmenting GPU topology. We present Spotlight, the first system that harvests spot GPUs for DiT RL post-training. Spotlight rests on two key insights we devise: (1)~we show that exploration can tolerate stale model weights because exploration that uses the model weights from the previous iteration preserves the relative ranking of random seeds, allowing exploration to run on idle spot GPUs during training. (2)~SP reconfiguration can reuse on-node state, reducing group recovery from minutes to sub-second launches. Built on these insights, Spotlight introduces three techniques: a bandit-based exploration planner that maximizes reward variance within the training time budget, elastic sequence parallelism that reconfigures SP groups on the fly via persistent schedulers and intra-node weight copying, and a preemption-aware pull-based request scheduler that balances load and commits in-flight state upon preemption. We implement Spotlight on the open-source RL platform ROLL and evaluate it on Qwen-Image post-training. Spotlight reaches the same target validation score $4\times$ faster than baselines, reducing total cost by $1.4$-$6.4\times$ while achieving superior image quality on DeepSeek-OCR and Geneval datasets with resolution $512\times512$ and $1280\times1280$.

20.
arXiv (CS.CV) 2026-06-17

ReAge3D: Re-Aging 3D Faces with View Consistency

We present a novel framework for realistic and controllable 3D face re-aging which produces highly detailed, identity-preserving results. Existing 3D editing methods, while effective for coarse semantic changes, are not well suited for re-aging, as even small inconsistencies across re-aged 2D views can lead to over-smoothing of subtle but perceptually important age-related details. To address this challenge, we first introduce a 2D diffusion-based re-aging model, DiffReaging, trained on synthetically generated image pairs. We further propose a center-out editing propagation strategy that leverages this re-aging model to reconstruct multi-view-consistent re-aged images. Specifically, starting from a re-aged frontal pivot view, we reconstruct the remaining views through warping and our proposed Masked-DiffReaging process. By injecting existing content at every step of the diffusion process, Masked-DiffReaging ensures that the reconstructed regions remain coherent with existing pixels. The resulting consistent set of re-aged views supervises the optimization of the re-aged 3D representation. Our method outperforms existing 3D editing techniques both visually and quantitatively, enabling smooth, fine-grained control over age transformations in 3D face models.

21.
arXiv (CS.CL) 2026-06-12

How reliable are LLMs when it comes to playing dice?

We investigate the probabilistic reasoning capabilities of large language models through a controlled benchmarking study on discrete probability problems. We constructed two datasets, respectively a set of standard exercises and a set of counterintuitive exercises, designed to trigger heuristic reasoning, and evaluated 8 state-of-the-art models, each tested with and without Chain-of-Thought prompting. Models achieve an average accuracy of 0.96 on standard problems but only 0.59 on counterintuitive ones. We further provide empirical evidence of token bias: performance drops by over 20% when canonical formulations are replaced by disguised variants. Embedding misleading suggestions in the prompt reduces performance by up to 34%, with no model proving immune. Taken together, the reported findings suggest that current LLMs are not yet genuine probabilistic reasoners, despite their success in advanced mathematical problems.

22.
arXiv (CS.CL) 2026-06-18

Lost in a Single Vector: Improving Long-Document Retrieval with Chunk Evidence Aggregation

Dense retrieval ranks one query vector against one document vector. On long documents, this interface can fail when a short but decisive span is weakened during document encoding before ranking. We study this failure mode as document-side early compression and introduce the Evidence Dilution Index (EDI) to measure how far a document-level representation falls below the strongest chunk-level evidence within the same gold document. Guided by this view, we propose DICE (Document Inference via Chunk Evidence), a training-free document-side strategy that splits documents into chunks, encodes them independently with a frozen model, and aggregates them back into a single vector while preserving the standard one-query-one-document interface. On LongEmbed, DICE improves retrieval across four backbones, with the largest gains on slices beyond 4k tokens: for Dream, Passkey >4k rises from 30.0 to 90.0 and Needle >4k from 23.3 to 74.0. Across 12,779 filtered samples, DICE yields lower EDI than the single-vector baseline in 92.8% of cases. These results establish document-level encoding as a practical and underexplored lever for long-document retrieval.

23.
arXiv (CS.CL) 2026-06-12

Reward Modeling for Multi-Agent Orchestration

Multi-Agent Systems (MAS) built on Large Language Models (LLMs) require effective orchestration to coordinate specialized agents, yet training such orchestrators is hindered by limited supervision and high computational cost. We propose Orchestration Reward Modeling (OrchRM), a self-supervised framework for evaluating orchestration quality without human annotations. OrchRM leverages intermediate artifacts from multi-agent executions to construct win-lose pairs for Bradley-Terry reward model training. Unlike existing MAS test-time scaling and orchestrator training frameworks that rely on costly sub-agent rollouts, OrchRM operates directly at the orchestration level, enabling efficient and high-performing reward-guided orchestrator training and MAS test-time scaling. OrchRM improves training efficiency by up to 10x in token usage while improving MAS test-time scaling performance by up to 8% in accuracy. These gains consistently transfer across multiple domains, including mathematical reasoning, web-based question answering, and multi-hop reasoning, demonstrating orchestration-level reward modeling as a scalable direction for robust multi-agent orchestration. Code will be available at https://github.com/Wang-ML-Lab/OrchRM.

24.
arXiv (CS.AI) 2026-06-11

Certifiable Safe RLHF: Semantic Grounding and Fixed Penalty Constraint Optimization for Safer LLM Alignment

arXiv:2510.03520v2 Announce Type: replace-cross Abstract: Ensuring safety is a foundational requirement for large language models (LLMs). Achieving an appropriate balance between enhancing the utility of model outputs and mitigating their potential for harm is a complex and persistent challenge. Contemporary approaches frequently formalize this problem within the framework of Constrained Markov Decision Processes (CMDPs) and employ established CMDP optimization techniques. However, these methods exhibit two notable limitations. First, their reliance on reward and cost functions renders performance highly sensitive to the underlying scoring mechanism, which must capture semantic meaning rather than being triggered by superficial keywords. Second, CMDP-based training entails tuning dual-variable, a process that is both computationally expensive and does not provide any provable safety guarantee for a fixed dual variable that can be exploitable through adversarial jailbreaks. To overcome these limitations, we introduce Certifiable Safe-RLHF (CS-RLHF) that introduces a cost model trained on a large-scale corpus to assign semantically grounded safety scores. In contrast to the lagrangian-based approach, CS-RLHF adopts a rectified penalty-based formulation. This design draws on the theory of exact penalty functions in constrained optimization, wherein constraint satisfaction is enforced directly through a suitably chosen penalty term. With an appropriately scaled penalty, feasibility of the safety constraints can be guaranteed at the optimizer, eliminating the need for dual-variable updates. Empirical evaluation demonstrates that CS-RLHF outperforms state-of-the-art LLM model responses rendering at-least 5 times efficient against nominal and jail-breaking prompts

25.
arXiv (CS.CL) 2026-06-15

Pragmatic Inference for Moral Reasoning Acquisition: Generalization via Metapragmatic Links

While moral reasoning has emerged as a promising research direction for large language models (LLMs), achieving robust generalization remains a critical challenge. This challenge arises from the gap between what is said and what is morally implied. In this paper, we build on metapragmatic links and Moral Foundations Theory to close this gap. Specifically, we develop a pragmatic inference approach that enables LLMs, given a moral situation, to acquire the metapragmatic links between moral reasoning objectives and the social variables that influence them. We adapt this approach to three different moral reasoning tasks to demonstrate its adaptability and generalizability. Experimental results show that our approach significantly enhances LLMs' generalization in moral reasoning, paving the way for future research to leverage pragmatic inference across a wide range of moral reasoning tasks.