# Research Papers — 2026-06 · bundle page 241 of 260 > Full-text markdown of papers #12001–12050 of 12993 published in 2026-06, newest first. AI-readable mirrors of the full 12993-paper 2026-06 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-06/page/241 [← Previous bundle](https://aipod.dev/papers/archive/2026-06/page/240.md) [Next bundle →](https://aipod.dev/papers/archive/2026-06/page/242.md) ### CoEval: Ranking Language Models for Custom Tasks Without Labeled Data or Trustworthy Benchmarks Authors: Alexander Apartsin, Yehudit Aperstein Published: 2026-06-02 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2606.03650v2 Selecting a pretrained language model, or evaluating a fine-tuned one, for a specific application is a high-value decision, yet the public benchmarks used to make it are poorly suited: a generic benchmark need not reflect a particular sub-domain or sub-task, and its scores are suspect when its items have leaked into pretraining and are recalled rather than solved. We present CoEval, an open framework that supplies a trustworthy, task-specific signal through ensemble self-evaluation: from a task or domain description, a pool of models rotates through all three roles, teacher, student, and judge, to generate a fresh, contamination-free benchmark, answer it, and score one another, with no human labels or raters. Because every model also answers as a student, the responses are the data that weight each question by its discriminative power and each judge by its consensus with the panel. Where ground truth exists, CoEval recovers the true ranking and tracks objective correctness at \r{ho}=0.86, and the weighting recovers the gold ranking of thirteen models at Spearman 0.95. Reliability comes from panel composition, not size: this label-free weighting zeroes out broken judges and down-weights saturated questions, so neither distorts the ranking. Generated items show zero verbatim overlap with five public benchmarks, the panel cancels verbosity bias and precludes same-family self-preference, and rankings are domain-specific: three different models top four de-novo domains, so a generic leaderboard misdirects most practitioners. The same pipeline reruns on each model release, giving any team a contamination-free leaderboard for its application. --- ### Safety Measurements for Fine-tuned LLMs Should be Grounded in Capability Authors: Krishnapriya Vishnubhotla, Hillary Dawkins, Isar Nejadgholi, Svetlana Kiritchenko Published: 2026-06-02 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2606.03648v1 Adapting foundation large language models to a user's task or preferred style through fine-tuning can result in compromising the model's safety. Previous works examined the effects of fine-tuning on model safety in limited and seemingly random experimental settings. We argue that anchoring fine-tuning to a specific capability goal is essential for avoiding arbitrary empirical choices, allowing us to draw meaningful conclusions about safety impacts, and to compare mitigation methods on a consistent basis. We conduct a multi-dimensional evaluation of the effects of fine-tuning on model behavior by focusing on capability as well as safety. Our results surface important issues that (1) fine-tuned models can produce incoherent generations in response to safety prompts, (2) automated safety judgments are unreliable for such incoherent outputs, and (3) the conclusions about the effects of fine-tuning can change depending on the choice of safety benchmark as well as the safety evaluator. --- ### Black-box, Adaptive, Efficient, Transferable, Harmful, Applicable... Attacks Are All You Need to Break LLMs Authors: Vincent Limbach, Jonas Dornbusch, David Lüdke, Stephan Günnemann, Leo Schwinn Published: 2026-06-02 Categories: cs.CR, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.03647v1 Accurately evaluating adversarial robustness is a longstanding challenge. A flawed attack design can inflate robustness estimates, making deployment risk assessment and defense comparison unreliable. Historically, standardized attacks such as AutoAttack have largely resolved this for image classifiers, providing a reliable evaluation baseline for systematic comparison across defenses. However, no equivalent exists for LLM jailbreak evaluation yet, where designing such an attack is considerably more difficult. A reliable attack must, among other things, be black-box compatible, applicable to arbitrary defense pipelines, and efficient, which no existing method jointly satisfies. We introduce Indirect Harm Optimization (IHO), a masked diffusion language model attacker trained via iterative preference optimization against a harmfulness judge, requiring only black-box access to the target. The same method can be used without modification as a strong adaptive attack on individual behaviors, or as an efficient amortized policy that transfers to held-out behaviors and unseen target models without fine-tuning. Even against layered defenses, such as a Circuit Breaker-trained model combined with an auxiliary detector, IHO improves attack success considerably over state-of-the-art approaches, without any defense-specific adaptation. Our results position IHO as a practical step toward the kind of standardized jailbreak evaluation that has improved reliability in the past. Code and models are available on GitHub and Hugging Face. --- ### A Benchmark for Semi-supervised Multi-modal Crowd Counting Authors: Haoliang Meng, Xiaopeng Hong, Yabin Wang, Wangmeng Zuo Published: 2026-06-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.03646v1 This paper constructs the first benchmark on semi-supervised multi-modal crowd counting. To lay the foundation for this unexplored task, we first formulate the semi-supervised multi-modal setting and a standardized protocol that specifies the labeled-unlabeled data partition across different labeled ratios. Next, to establish solid reference points, we carefully tailor a diverse set of representative baselines, including existing fully supervised multi-modal methods and semi-supervised single-modal methods. Then, we carefully evaluate their performance under our proposed benchmark. Codes and the data partition will be released on https://github.com/HenryCilence/Semi-supervised-Multimodal-Crowd-Counting. --- ### Gender-Dependent Diagnostic Substitution in LLM Medical Triage: Same Symptoms, Unequal Urgency Authors: Qi Han Wong Published: 2026-06-02 Categories: cs.AI, cs.CY arXiv: https://arxiv.org/abs/2606.03641v1 We investigate whether large language models produce different medical triage recommendations for identical neurological symptoms when only the patient's stated gender and age vary. Using three model families--Gemini 3.5 Flash, Claude Sonnet 4.6, and GPT-5.4-mini--we present a standardized symptom profile (persistent headache, blurred vision, morning nausea, visual disturbances) across seven demographic conditions: three age groups (25, 38, 65) x two genders (male, female), plus a gender-unspecified baseline (n = 30 per condition per model, 630 total trials). We find a stark, systemic gender-dependent triage disparity: young women receive significantly lower emergency room (ER) referral rates than age-matched men (Gemini: 0% vs. 23.3%; Claude: 6.7% vs. 96.7%; GPT: 6.7% vs. 66.7%, all p < 0.001). The disparity disappears at age 65 for all models. The primary mechanism is diagnostic substitution: the models anchor on a gender-associated diagnosis, preferentially classifying young women with Idiopathic Intracranial Hypertension (IIH)--a condition epidemiologically linked to women of childbearing age--while diagnosing men with generic increased intracranial pressure with space-occupying lesions in the differential. This diagnostic closure routes female patients to lower-urgency care (outpatient doctor appointments) despite comparable severity ratings (7-9/10). Our findings demonstrate that clinical LLMs replicate documented human clinical biases by using epidemiological priors to suppress triage urgency, suggesting that AI triage engines must decouple urgency assessment from probabilistic diagnostic priors. We release all code, prompts, and raw results. --- ### VidMsg: A Benchmark for Implicit Message Inference in Short Videos Authors: Issar Tzachor, Michael Green, Rami Ben-Ari Published: 2026-06-02 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.03635v1 Understanding short online videos involves more than identifying visible objects and actions; video makers often include an underlying message or purpose in the clip. We introduce VidMsg, a benchmark for evaluating implicit message understanding in short, internet-native video clips. VidMsg contains 400 YouTube-derived clips across 9 practical topic areas and 52 fine-grained target messages, covering domains such as career and finance, education, health and well-being, culture, safety, sustainability, and lifestyle. VidMsg is constructed through a message-first pipeline: an LLM first translates target messages into indirect search scenarios, which are used to retrieve candidate clips. Human annotators then retain clips that convey the intended message without being overly explicit. VidMsg is designed primarily for bidirectional message-clip retrieval for scalable applications such as video search and recommendation, where systems must capture holistic video understanding. In addition to retrieval, VidMsg includes a diagnostic multiple-choice QA benchmark, where models select the intended message of a clip from semantically related alternatives. Experiments with contemporary video-language and retrieval models show that strong models often fail on VidMsg, because the task requires pragmatic inference, integration of contextual cues, and discrimination among semantically close messages. We also introduce VidVec-Msg, a baseline method that improves message-oriented retrieval while leaving substantial headroom for future work. --- ### AnchorMoE: Interpretable Time Series Classification via Anchor-Routed MoE Authors: Tao Xie, Zexi Tan, Haoyi Xiao, Mengke Li, Yiqun Zhang, Yang Lu, Cuie Yang, Yiu-ming Cheung Published: 2026-06-02 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.03631v3 Multivariate time series classification (MTSC) is pivotal in high-stakes domains, such as clinical diagnosis and industrial fault detection, where safe deployment necessitates transparent decision-making. However, isolating the temporal segments that drive model predictions is challenging because discriminative signals in real-world time series are typically sparse, heterogeneous, and heavily obscured by background noise. This paper, therefore, proposes AnchorMoE, an interpretable-by-construction classification framework. Built upon a Mixture-of-Experts (MoE) architecture, AnchorMoE encodes multi-view representations of local patches and routes them to specialized experts, ensuring that the final prediction is formulated as an exact additive decomposition over the input segments, facilitating ante-hoc transparency rather than relying on post-hoc estimations. To maintain the reliability of this decomposition under sparse signal distributions, we introduce a geometric orthogonality constraint that penalizes representational redundancy, compelling distinct experts to specialize in heterogeneous predictive patterns. Furthermore, an uncertainty-aware reliability gate is designed to dynamically calibrate the contribution of each segment, effectively suppressing residual background noise. Extensive experiments on real-world and synthetic benchmarks demonstrate that AnchorMoE achieves highly competitive classification performance while faithfully grounding its decisions in the raw time series. --- ### TSQAgent: Rating Time Series Data Quality via Dedicated Agentic Reasoning Authors: Shunyu Wu, Dan Li, Haozheng Ye, Weibin Feng, Jian Lou, Bo Zhang, Wenjie Feng, Chenjuan Guo, See-Kiong Ng Published: 2026-06-02 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.03629v1 Assessing the quality of time series (TS) data is fundamental yet inherently challenging due to the multifaceted nature of quality dimensions. Recently, large language models (LLMs) have emerged as a promising paradigm for TS quality assessment via pairwise comparison and per-dimension evaluation. However, existing approaches rely on manually predefined quality dimensions and purely text-based reasoning, leaving it unknown whether LLMs can identify truly relevant quality dimensions or perform grounded and quantitative quality comparisons. To investigate this, we construct TSQBench, a dedicated benchmark for evaluating LLMs on two progressive capabilities: (i) understanding and identifying relevant quality dimensions, and (ii) performing quality comparison under specific dimensions. Our analysis reveals that current LLMs consistently struggle with both dimension identification and evidence-grounded quality comparison. To address these limitations, we propose TSQAgent, a novel agentic reasoning framework for TS quality rating consisting of three collaborative roles: Perceiver for focused dimension selection, Inspector for dimension-wise quantitative analysis, and Adjudicator that aggregates and refines the final judgment. In particular, we introduce an agentic reasoning strategy that instills the ability to identify and prioritize the most relevant quality dimensions, and further propose an agent workflow equipped with external analytical tools to enable precise quantitative comparisons over selected dimensions. Experiments on both the proposed benchmark and eleven real-world datasets demonstrate that our framework not only substantially improves LLMs' capabilities in quality understanding and quantitative comparison but also effectively translates these improvements into better quality-aware data selection, leading to enhanced downstream performance and data efficiency. --- ### Building Reliable Long-Form Generation via Hallucination Rejection Sampling Authors: Lin Li, Georgia Channing, Suhaas M Bhat, Gabriel Davis Jones, Yarin Gal Published: 2026-06-02 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.03628v1 Large language models (LLMs) have achieved remarkable progress in open-ended text generation, yet they remain prone to hallucinating incorrect or unsupported content, which undermines their reliability. This issue is exacerbated in long-form generation due to hallucination snowballing, a phenomenon where early errors propagate and compound into subsequent outputs. To address this challenge, we propose a novel inference-time hallucination mitigation framework, named Segment-wise HAllucination Rejection Sampling (SHARS), which uses an arbitrary hallucination detector to identify and reject hallucinated segments during generation and resample until faithful content is produced. By retaining only confident information and building subsequent generations upon it, the framework mitigates hallucination accumulation and enhances factual consistency. To instantiate this framework, we adopt semantic uncertainty as the detector and introduce several vital modifications to address its limitations and better adapt it to long-form text. Our method enables models to self-correct hallucinations without requiring external resources such as web search or knowledge bases, while remaining compatible with them for future extensions. Empirical evaluations on standardized hallucination benchmarks demonstrate that our method substantially reduces hallucinations in long-form generation while preserving or even improving the informativeness of generation. Code is available at: https://github.com/TreeLLi/hallucination-rejection-sampling. --- ### TurtleAI: Benchmarking Multimodal Models for Visual Programming in Turtle Graphics Authors: Chao Wen, Jacqueline Staub, Adish Singla Published: 2026-06-02 Categories: cs.CV, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2606.03626v1 Vision-language models (VLMs) have been explored for visual programming, where they generate code to solve visual tasks. However, most prior work focuses on visual programming for productivity; it remains unclear how well current VLMs perform on education-oriented visual programming and what factors limit their performance. To bridge this gap, we introduce TurtleAI, a benchmark containing 823 tasks curated based on real-world visual programming tasks in the Turtle Graphics domain. Solving these tasks requires models to perceive geometric patterns, reason about spatial relationships, and synthesize Python code that faithfully reproduces geometric patterns. We evaluate 20+ VLMs, including GPT-5, GPT-4o, and Qwen2-VL-72B, and find that they struggle significantly, with most achieving success rates below 30%. To address these limitations, we propose a data generation technique that requires only a small set of seed samples. Fine-tuning Qwen2-VL-72B on the resulting synthetic data yields an improvement of about 20% on real-world tasks. Our failure analysis reveals that GPT-4o struggles with spatial reasoning and precise visual replication, whereas fine-tuning primarily improves the alignment between visual reasoning and code implementation. --- ### Bridging Auxiliary Constraints to Resolve Instruction Following in Large Reasoning Models Authors: Zhengyi Zhao, Shubo Zhang, Huimin Wang, Zezhong Wang, Yutian Zhao, Yefeng Zheng, Binyang Li, Yulan He, Kam-Fai Wong, Xian Wu Published: 2026-06-02 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2606.03624v1 Large Reasoning Models (LRMs) have demonstrated impressive capabilities in many tasks, yet they struggle with reliably following multiple instructions, either by failing to satisfy individual constraints or by struggling to balance competing constraints simultaneously. We formalize this challenge as the Constraint Adherence Problem (CAP). This paper introduces a novel framework that addresses CAP by representing instructions as a structured knowledge graph of constraints. Our approach, Constraint Relationship Graph Completion (CRGC), explicitly models relationships between constraints, identifies adherence challenges, and discovers ``bridge constraints'' that help the model better focus on and reconcile requirements. Bridge constraints act as auxiliary instructions that make primary constraints more salient and compatible. Unlike existing approaches that enhance instruction following through general training methods, CRGC specifically improves constraint satisfaction by leveraging the model's own knowledge to create better pathways for generation. Experiments across three popular instruction following datasets demonstrate that our approach reduces constraint violations by 39% compared to standard prompting while maintaining reasoning abilities of large reasoning models. --- ### Physics-Guided Policy Optimization with Self-Distillation Authors: Ke Wang, Yuning Wu, Haoran Liu, Chaoqun Jia, Devin Chen, Kai Wei Published: 2026-06-02 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.03620v1 Self-distilled policy optimization (SDPO) has become a popular paradigm for LLM post-training, where a model learns from its own predictions conditioned on privileged information. SDPO, however, is sensitive to how much each update step should be trusted: corrections from a self-teacher can be highly informative on some batches and misleading on others, and applying them uniformly with a fixed step size can destabilize training. Drawing inspiration from viscous-fluid dynamics and formalizing the analogy at the SDE level, we propose Physics-Guided Policy Optimization (PGPO), which introduces an information-modulated step-size multiplier derived from a mutual-information estimate between the student's predictions and the feedback-conditioned teacher. We show that this modulation preserves the order-1 weak-approximation guarantees of vanilla SGD, and incurs negligible overhead per iteration. We evaluate PGPO on the Science-QA dataset, where it outperforms SDPO on 3 of the 4 domains with gains of up to +4.5 points, while remaining stable in a setting where SDPO collapses late in training. --- ### Cross-Lingual Token Arbitrage: Optimizing Code Agent Context Windows via Local LLM Preprocessing Authors: Mehmet Utku Colak Published: 2026-06-02 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.03618v2 AI-assisted coding agents are bottlenecked by input-token cost. Two pathologies of raw human input drive much of this overhead: tokenization inefficiency for non-English text and structural entropy in conversational prompts. Existing approaches act reactively by compressing already-bloated contexts or intervening after failures occur. We introduce a pre-flight, edge-side prompt-rewriting middleware that operates between the developer and the cloud agent. A local Llama 3.2 (3B) model performs cross-lingual translation into English, structural rewriting into a compact task-oriented format, and regex-validated rewrite-with-fallback safeguards to ensure the optimized prompt is never larger than the original. We evaluate on OMH-Polyglot, a multilingual coding benchmark spanning Turkish, Arabic, Chinese, and code-switched specifications. Across three commercial LLM backends, the middleware reduces prompt tokens by 34-47 percent and total tokens by up to 18.8 percent while preserving or improving task accuracy. Ablation studies show that gains arise primarily from the rewriting stage rather than simple function-name extraction. Compared with LLMLingua-2 at matched compression rates, our method consistently achieves superior OckScore performance across all evaluated backends. These results demonstrate that proactive prompt optimization can substantially reduce inference costs without sacrificing coding quality. --- ### SkelHCC: A Hyperbolic CLIP-Driven Cache Adaptation Framework for Skeleton-based One-Shot Action Recognition Authors: Yanan Liu, Anqi Zhu, Jingmin Zhu, Jun Liu, Hossein Rahmani, Mohammed Bennamoun, Farid Boussaid, Dan Xu, Qiuhong Ke Published: 2026-06-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.03610v1 Skeleton-based action recognition aims to understand human behaviors from body joint sequences and is especially challenging in the one-shot setting, where only a single labeled exemplar is available for each novel action. A key challenge is learning representations that capture the hierarchical and compositional structure of human motion while aligning effectively with high-level action semantics under extreme data scarcity. Existing approaches, largely based on Euclidean embeddings and low-level motion cues, struggle to model the tree-like organization of skeleton data, limiting cross-modal alignment and generalization to unseen action categories. We propose SkelHCC, a unified skeleton hyperbolic CLIP-driven cache adaptation framework for one-shot skeleton-based action recognition. SkelHCC introduces an Explicitly Hierarchical Hyperbolic CLIP (EH-HCLIP) module that embeds skeleton sequences and action language into a shared hyperbolic space. By leveraging the negative curvature and exponential volume growth of hyperbolic geometry, EH-HCLIP naturally encodes the joint-part-body hierarchy of human anatomy and yields structurally consistent cross-modal representations. To support efficient one-shot adaptation, SkelHCC further integrates a training-free LLM-guided Multi-granularity Voting Cache (LMV-Cache) for context-aware inference. Experiments on NTU RGB+D 60, NTU RGB+D 120, and PKU-MMD demonstrate that SkelHCC consistently outperforms state-of-the-art methods. --- ### A 3D Isovist World Model -- Revealing a City's Unseen Geometry and Its Emergent Cross-City Signature Authors: Xuhui Lin, Stephen Law, Nanjiang Chen, Kunyao Li, Tao Yang Published: 2026-06-02 Categories: cs.RO, cs.LG arXiv: https://arxiv.org/abs/2606.03609v3 Embodied agents that navigate cities rely on world models that predict how their surroundings will change as they move. But for navigation, what matters is not what the buildings look like; it is where the agent can go. Most world models nonetheless predict appearance, learning how a scene looks rather than the space an agent can move through. Those that do target geometry, such as bird's-eye-view occupancy grids, flatten the three-dimensional environment onto a ground plane, discarding the above-ground and multi-level structure that shapes real navigation. What is missing is a predictive target that captures the navigable geometry an agent actually traverses, without photometric entanglement and without collapsing the third dimension. Our key idea is to model the open volume between buildings, the negative space, encoded as a 3D isovist: a spherical visibility-depth map recording the distance to the nearest surface in every direction. We introduce an embodied world model that predicts the next isovist from a short history of past isovists and a movement action. The prediction is formulated as a depth residual so the decoder inherits sharp building edges, trained with self-rollout scheduled sampling to keep corrupted context on the geometry manifold, and equipped with a persistent latent bird's-eye-view spatial map for cross-path consistency. Our central finding is emergent and unexpected: a single city-blind model trained on Manhattan and Paris develops a cross-city spatial signature, with city identity linearly decodable from its temporal latents far above single-frame baselines, so the signature lives in the learned dynamics rather than in appearance. The representation is lightweight, interpretable, and reproducible, offering a geometric substrate for spatial reasoning in embodied AI, robotics, and urban analysis, released with an open dataset and pipeline. --- ### Exploiting Verification-Generation Gap: Test-Time Reinforcement Learning with Confidence-Conditioned Verification Authors: Jiahui Li, Jianfeng Shan, Wenpei Chen, Shunyu Wu, Jian Lou, Wenjie Feng, Dan Li, See-Kiong Ng Published: 2026-06-02 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.03608v1 Test-time reinforcement learning has emerged as a promising paradigm for enhancing the complex reasoning abilities of large language models in a completely label-free manner. Despite existing studies focusing on Pass@1 performance, optimizing Pass@k remains under-explored yet critical in label-free settings, which measures generation coverage for sustained exploration. Optimizing Pass@k in label-free setting is highly non-trivial, as directly applying the Pass@k advantage designs effective for RLVR yields unsatisfactory performance. Through in-depth empirical analysis, we discover the root causes hindering performance: pseudo-label estimations for low-confidence samples have a high probability of being incorrect, while candidate answers for high-confidence samples suffer from severe diversity collapse. To overcome these hurdles, we propose TTRL-CoCoV (Test-Time Reinforcement Learning with Confidence-Conditioned Verification), a novel confidence-adaptive framework that expands Pass@k coverage and improves Pass@1 performance. Based on our key insight that verification capability generally leads generation capability, TTRL-CoCoV employs a confidence-conditioned mechanism: for high-confidence samples, it bootstraps verifier and applies an exploration-enhancing reward to prevent diversity collapse; for low-confidence samples, it delegates pseudo-label selection to the verifier to filter incorrect pseudo-labels; and for medium-confidence samples, it bypasses verification entirely. Extensive experiments demonstrate that TTRL-CoCoV outperforms the best competing methods across 6 widely-recognized benchmarks, achieves average absolute gains of +9.8% in Pass@1 and +18.7% in Pass@16 over TTRL, and even achieves absolute Pass@1 improvements of up to +5.0% across multiple reasoning benchmarks when compared against fully supervised RL methods. Our code repository: https://github.com/shanjf666/CoCoV. --- ### Testing LLM Arithmetic Reasoning Generalization with Automatic Numeric-Remapping Attacks Authors: Malia Barker, Bishal Lakha, Edoardo Serra, Francesco Gullo Published: 2026-06-02 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2606.03606v2 Large language models achieve strong performance on arithmetic reasoning benchmarks, and one common response to arithmetic brittleness is to delegate computation to code. Yet models are still often used in settings where they must reason directly from natural language, and trustworthy models should solve small-number arithmetic word problems without external tools. Prior work shows that LLMs are sensitive to numerical variation: a model may solve an original problem but fail on structurally similar variants requiring the same reasoning procedure with different numbers. We ask whether this fragility persists under a stricter setting involving small, schema-preserving numeric changes that retain the original reasoning program and avoid large-number stress tests. We introduce an automatic algorithm for generating numeric-remapping attacks on arithmetic word problems. Unlike template-based perturbation methods requiring manual schemas or constraints, our approach derives problem-specific symbolic representations, generates constrained numeric remappings, recomputes gold answers, and realizes transformed questions through deterministic edits guided by LLM-generated edit plans. Stage-wise validation and a high-confidence audit retain reliable attacks, making the pipeline scalable with limited human intervention. We evaluate DeepSeek-R1 (70B), Gemma4 (31B), and GPT-OSS (120B) on GSM8K, MAWPS, and MultiArith. On GSM8K, completed runs show conditional accuracy drops of 12.16 to 25.82 percentage points. MAWPS and MultiArith are far more stable, with most attacked accuracies near or above 98%. These results show that numeric-remapping robustness depends strongly on dataset structure: GSM8K remains sensitive even when reasoning programs are preserved and answers are recomputed, while shorter, more regular datasets are more robust. --- ### Beyond the Literal: Decomposing Pragmatic Intent in Multimodal Meme Understanding Authors: Zhengyi Zhao, Shubo Zhang, Zezhong Wang, Luyao Ye, Huimin Wang, Hanqi Yan, Binyang Li, Kam-Fai Wong, Yulan He Published: 2026-06-02 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.03604v1 When asked what a meme or sarcastic post means, Large Vision Language Models (LVLMs) tend to describe what the image shows rather than what the author is trying to communicate. Standard instruction tuning entangles a post's literal content with its pragmatic meaning, letting surface-level details contaminate the final response. We reframe meme understanding as a problem of literal-pragmatic decomposition and propose \textbf{Intent Projection}, a framework that separates the two signals at the representation, output, and objective levels within a single LVLM backbone. At the representation level, an orthogonal projection module removes dominant unimodal directions from the fused image-text representation, retaining only the pragmatic residual, while a surface-real affect classifier anchors the decoder with a discrete tag that names the polarity gap. At the output level, the model externalizes a structured reasoning chain, and at the objective level a contrastive reward explicitly penalizes answers that restate the literal description. Across six multimodal benchmarks, Intent Projection consistently outperforms open-source baselines and narrows the gap to proprietary models, with the largest gains on high-divergence posts where literal collapse is most damaging. --- ### World Models Meet Language Models: On the Complementarity of Concrete and Abstract Reasoning Authors: Yucheng Zhou, Wei Tao, Yiwen Guo, Jianbing Shen Published: 2026-06-02 Categories: cs.CV, cs.CL arXiv: https://arxiv.org/abs/2606.03603v2 World models and multimodal large language models (MLLMs) provide complementary capabilities for predicting future outcomes from static visual observations. World models can generate concrete visual rollouts of possible futures, while MLLMs can reason abstractly over questions, goals, and rules. However, generated rollouts are stochastic and may be visually plausible but task-incorrect, making it necessary to determine when visual simulation is useful, whether a rollout is credible, and how it should influence the final answer. We formulate this problem as controlled concrete reasoning, where a model learns to invoke, verify, and integrate visual future simulation alongside abstract reasoning. To study this setting, we construct two human-verified benchmarks, VRQABench for controllable spatial lookahead and OpenWorldQA for open-domain physical prediction, and propose Privileged-Future On-Policy Self-Distillation (PF-OPSD). During training, PF-OPSD uses ground-truth future videos and answers only as teacher-side privileged context to evaluate on-policy concrete-reasoning trajectories, while the deployable student never observes true futures at test time. Experimental results show that PF-OPSD outperforms baseline by 10.6% and 10.9% on VRQABench and OpenWorldQA, respectively, while increasing robustness to noisy or conflicting rollouts. Our code and dataset are available at https://github.com/yczhou001/PF-OPSD. --- ### World Models Meet Language Models: On the Complementarity of Concrete and Abstract Reasoning Authors: Yucheng Zhou, Wei Tao, Yiwen Guo, Jianbing Shen Published: 2026-06-02 Categories: cs.CV, cs.CL arXiv: https://arxiv.org/abs/2606.03603v1 World models and multimodal large language models (MLLMs) provide complementary capabilities for predicting future outcomes from static visual observations. World models can generate concrete visual rollouts of possible futures, while MLLMs can reason abstractly over questions, goals, and rules. However, generated rollouts are stochastic and may be visually plausible but task-incorrect, making it necessary to determine when visual simulation is useful, whether a rollout is credible, and how it should influence the final answer. We formulate this problem as controlled concrete reasoning, where a model learns to invoke, verify, and integrate visual future simulation alongside abstract reasoning. To study this setting, we construct two human-verified benchmarks, VRQABench for controllable spatial lookahead and OpenWorldQA for open-domain physical prediction, and propose Privileged-Future On-Policy Self-Distillation (PF-OPSD). During training, PF-OPSD uses ground-truth future videos and answers only as teacher-side privileged context to evaluate on-policy concrete-reasoning trajectories, while the deployable student never observes true futures at test time. Experimental results show that PF-OPSD outperforms baseline by 10.6% and 10.9% on VRQABench and OpenWorldQA, respectively, while increasing robustness to noisy or conflicting rollouts. Our code and dataset are available at https://github.com/yczhou001/PF-OPSD. --- ### CauTion: Knowing When to Trust LLMs for Ensemble Causal Discovery Authors: Bo Peng, Kaiwen Wu, Sirui Chen, Zhiheng Wang, Yu Qiao, Chaochao Lu Published: 2026-06-02 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2606.03602v1 Causal discovery from observational data remains challenging due to the fundamental limitations of purely statistical methods, such as statistical distinguishability within equivalence classes and sensitivity to finite sample sizes. While large language models (LLMs) offer a promising source of domain knowledge to complement statistical inference, existing LLM-augmented methods are vulnerable to LLM errors and incur high token costs. Moreover, reliance on a single data-centric algorithm can make results sensitive to algorithm-specific biases. To address these limitations, we propose CauTion, a framework that reliably integrates LLM domain knowledge into an ensemble of statistical causal discovery algorithms through consensus filtering and LLM reliability estimation. CauTion proceeds in three stages. First, an algorithm ensemble utilizes a consensus voting to resolve up to 96% of edges on which algorithms agree, achieving near-perfect accuracy on the filtered consensus edges. Second, a trust-calibrated arbitration mechanism estimates the relative reliability of the LLM and the algorithms via an annotation-free trust calibration procedure, which is then utilized to govern a trust-weighted voting process that restricts LLM arbitration exclusively to edges with unreliable algorithmic evidence. Third, a cycle repair step is applied to guarantee the final causal graph is validly acyclic. Experiments on six datasets demonstrate that CauTion consistently outperforms both data-centric and LLM-augmented baselines, with larger gains on larger graphs and strong robustness to LLM errors. Code is available at https://github.com/OpenCausaLab/CauTion. --- ### DDOR: Delta Debugging for Explainable Overrefusal Testing and Repair Authors: Qinyan Zhou, Peixin Zhang, Jun Sun, Haonan Zhang, Dongxia Wang Published: 2026-06-02 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2606.03601v1 While safety alignment and guardrails help large language models (LLMs) avoid harmful outputs, they can also induce overrefusal, i.e., unwarranted rejection of benign queries that merely appear risky. We present DDOR (Delta Debugging for OverRefusal), a fully automated and explainable framework for overrefusal testing and repair in a black-box setting, where only model inputs and outputs are accessible and internal safety mechanisms remain opaque. DDOR applies delta debugging to localize minimal refusal-triggering fragments (mRTFs) that provide phrase-level, explainable evidence for why a refusal occurs. Conditioned on these mRTFs, DDOR generates diverse, context-rich prompts and performs multi-oracle validation to filter intrinsically unsafe or ambiguous cases, producing scalable and model-specific overrefusal test suites (approximately 1K cases per model). Beyond evaluation, we further leverage localized mRTFs to perform targeted prompt repair, substantially reducing overrefusal while preserving the original intent and maintaining safety on genuinely harmful inputs. Overall, DDOR offers a practical end-to-end solution to both evaluate and mitigate overrefusal, improving LLM usability without sacrificing safety. --- ### Set-Preserving Calibration from Conformal P-Values to E-Values Authors: Nabil Alami, Jad Zakharia, Souhaib Ben Taieb Published: 2026-06-02 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2606.03600v1 Standard conformal prediction (CP) procedures are typically formulated in terms of p-values, but reliance on p-values alone limits flexibility, for example, when combining dependent evidence across models or data splits. Recent work has explored e-value formulations for conformal inference, yet a direct connection between p- and e-value formulations in CP has been missing, especially regarding their statistical efficiency. We first identify limitations of classical p-to-e calibrators in the CP setting, showing that they are not set-preserving and can lead to overly conservative prediction sets. To address this, we propose a novel P2E calibrator that converts conformal p-values into e-values without altering the prediction set induced by the original conformal p-value. We establish both theoretically and empirically that our calibrator can yield significant efficiency gains over existing p-to-e calibrators. This e-value formulation enables principled use of recent advances in e-value merging and randomization, where we demonstrate its impact in two applications: cross-conformal prediction (CCP), whose variants typically provide only approximate $1-2α$ coverage, and conformal aggregation (CA). In both cases, our e-value-based methods satisfy the desired $1-α$ coverage guarantee while improving efficiency over standard baselines. More broadly, our approach expands the flexibility of CP and opens new directions for efficient, distribution-free uncertainty quantification. --- ### PHASER: Phase-Aware and Semantic Experience Replay for Vision-Language-Action Models Authors: Ziyang Chen, Shaoguang Wang, Weiyu Guo, Qianyi Cai, He Zhang, Pengteng Li, Yiren Zhao, Yandong Guo Published: 2026-06-02 Categories: cs.RO, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2606.03598v2 Vision-Language-Action (VLA) models have achieved remarkable success in language-conditioned robotic manipulation. However, deploying these models in open-ended environments requires continuously acquiring novel skills, a process that inevitably triggers severe catastrophic forgetting of previously learned behaviors. While experience replay (ER) serves as a standard mitigating strategy, naive uniform sampling fundamentally misaligns with the temporal characteristics of manipulation trajectories. It systematically under-samples brief but causally critical sub-skills, leading to phase starvation, and completely overlooks the varying degrees of forgetting across historical tasks. To overcome these limitations, we introduce PHASER, an architecture-agnostic continual learning framework. PHASER employs a phase-centric capacity allocation to guarantee equal memory support for all sub-skills, coupled with a multi-modal interference routing strategy that dynamically prioritizes historical phases at high risk of forgetting. Furthermore, to enable fully autonomous lifelong adaptation, we integrate Auto-PC, a lightweight pipeline combining unsupervised action-signal change-point detection with VLM-based semantic verification to extract temporal boundaries without intensive manual supervision. Evaluated across three VLA backbones on LIBERO continual learning suites, PHASER yields substantial empirical improvements, increasing Average Success Rate (ASR) by up to 31% over matched-budget ER and achieving an 87.8% final ASR on the LIBERO-Goal CL setting. --- ### Multimodal Transformer Based Generic Mixture Density Network for Scattering Timescale Estimation of Fast Radio Bursts Authors: Bikash Kharel, Emmanuel Fonseca, Srinjoy Das, Mason Ng, Paul Scholz, Mawson W. Simmons, Lordrick Kahinga, Afrokk Khan Published: 2026-06-02 Categories: astro-ph.HE, stat.ML arXiv: https://arxiv.org/abs/2606.03596v1 The discovery rate of fast radio bursts (FRBs) continues to increase with the advent of new radio facilities and yet extracting their astrophysical parameters such as scattering timescale ($τ$) remains a significant bottleneck. Current $τ$ measurement approaches like fitting analytic template models and scattering aware de-convolution are accurate but slow, sensitive to initialization, limited by low signal to noise and often require manual supervision. These limitations inspired us to explore fast, robust and scalable machine learning methods to estimate the astrophysical parameter value. We present a deep learning approach named Multimodal Transformer Based Generic Mixture Density Network (MT-GMDN) which ingests FRB dynamic spectrum and its corresponding timeseries profile through parallel transformer encoders, fuses their latent representations and predicts the distribution of $τ$ with probabilistic output derived from generic mixture-density formulation. This formulation not only estimates the value of $τ$ but also captures the (zero inflated) nature of FRB populations where a significant fraction of bursts exhibit unresolvable scattering. We trained MT-GMDN on $\sim3500$ FRBs from CHIME/FRB \cattwo while holding out some fraction of FRBs for validation during training and for testing after the training completes. The model achieves a coefficient of determination ($R^2$) value of $94\%$ on the expected value of $τ$ for the events with measurable scattering with an excellent recall value of $90\%$ on the test data set. The model was also able to incorporate heteroskedastic errors enabling us the construction of a confidence interval for the predictions. --- ### Training a Predictive Coding Network on ImageNet using Equilibrium Propagation Authors: Tugdual Kerjan, Rasmus Høier, Benjamin Scellier Published: 2026-06-02 Categories: cs.LG, cond-mat.dis-nn, cs.NE arXiv: https://arxiv.org/abs/2606.03584v1 Equilibrium Propagation (EP) is a physics-based training framework that has primarily been employed in energy-based models, including continuous Hopfield networks, nonlinear resistive networks and coupled phase oscillators. However, EP's practical applications have so far remained limited to relatively small-scale problems. Predictive coding networks (PCNs), another class of energy-based models rooted in computational neuroscience, are typically trained with a specialized algorithm and have likewise not yet been demonstrated at large scale. In this work, we develop an EP-based training method for PCNs which combines the centered variant of EP with a novel equilibration scheme for PCNs. Using this approach, we train a 10-layer convolutional PCN (VGG10) on full-size ImageNet, achieving 13.23\% test error rate on the top-5 classification task, close to the 12.2\% backpropagation baseline. To our knowledge, this is the first demonstration of both PCNs and EP-based training at ImageNet scale. These results significantly extend the scalability of both approaches and suggest that the primary challenges in scaling EP in other physical systems may come more from the computational properties of these systems than from inherent limitations of the EP framework. --- ### UnsOcc: 3D Semantic Occupancy Prediction in Unstructured Scene via Rendering Fusion Authors: Ye Wu, Ruiqi Song, Baiyong Ding, Nanxin Zeng, Junjie Cheng, Yunfeng Ai Published: 2026-06-02 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2606.03581v1 Unstructured scenes present unique challenges for autonomous driving, as irregular obstacles and sparse scene layouts undermine the effectiveness of traditional perception methods such as 3D object detection. 3D semantic occupancy prediction has emerged as a prominent focus due to its ability to provide dense spatial representations by assigning semantic labels to individual voxels in 3D space. However, directly applying 3D semantic occupancy prediction to unstructured scenes remains challenging because scene sparsity hinders effective cross-modal fusion and the more severe long-tail distribution in these scenarios further degrades prediction performance. To validate the effectiveness of our approach, we construct a dedicated dataset of unstructured scenes collected from open-pit mines. Based on this, we propose UnsOcc, a multi-modal 3D semantic occupancy prediction framework that improves robustness in unstructured environments. At its core, we introduce a rendering-based fusion module, RenderFusion, which enhances cross-modal feature alignment through bidirectional rendering supervision. Furthermore, we propose GSRefinement, a detail-aware auxiliary supervision method based on Gaussian Splatting that projects sparse 3D occupancy predictions into dense 2D semantic segmentation maps, enabling effective supervision for long-tail categories. Extensive experiments on both the open-pit mine dataset and the nuScenes dataset demonstrate that our method significantly outperforms existing state-of-the-art approaches. --- ### Diffusing in the Right Space: A Systematic Study of Latent Diffusability Authors: Tianxiong Zhong, Xingye Tian, Xuebo Wang, Xin Tao, Pengfei Wan Published: 2026-06-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.03578v1 Latent diffusion models leverage visual tokenizers to compress images into latent spaces for efficient generative modeling. However, better reconstruction quality of a tokenizer does not necessarily translate into better generation quality, suggesting that latent representations should be evaluated not only by fidelity but also by their diffusability. Recent studies have proposed diverse explanations for diffusion-friendly latent spaces, including semantic separability, affine equivariance, distribution uniformity, spatial structure, spectral smoothness, and manifold continuity. Yet these properties are often validated on a limited set of tokenizers, leaving it unclear which factors are most predictive of downstream generation quality and whether such conclusions hold beyond the specific settings in which they are introduced. In this work, we conduct a systematic study of latent diffusability by training a large collection of tokenizers with diverse regularization strategies, architectures, and latent configurations, and evaluating them with multiple downstream diffusion backbones. Our analysis identifies several latent properties that consistently correlate with generation quality and exhibit strong generalization across experimental settings. Beyond existing metrics, we introduce Velocity Irreducible Variance (VIV), a measure of velocity ambiguity induced by trajectory crossings. Extensive experiments show that VIV is one of the most stable predictors of generation quality. --- ### Eliciting Complex Spatial Reasoning in MLLMs through Wide-Baseline Matching Authors: Hao Zhong, Muzhi Zhu, Shenyan Zeng, Anzhou Li, Cong Chen, Hua Geng, Duochao Shi, Wentao Ye, Tao Lin, Hao Chen, Chunhua Shen Published: 2026-06-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.03577v1 Wide-baseline matching (WBM) requires integrating geometric understanding, viewpoint changes, fine-grained perception, and occlusion reasoning, making it a challenging testbed for spatial reasoning in multimodal large language models (MLLMs) deployed in physical environments. However, current MLLMs lack systematic evaluation and training frameworks for these capabilities. We introduce ReasonMatch-Bench, a benchmark stratified by viewpoint displacement and matching granularity across indoor, outdoor, and object-centric scenarios, and show that current MLLMs still struggle with fine-grained wide-baseline correspondence: on a difficult 90-sample subset, human annotators achieve 84.0 F1, while the best existing baseline reaches 37.2. To bridge this gap, we build a scalable data-generation pipeline that automatically extracts wide-baseline view pairs from large-scale video-3D corpora, including RGB-D videos and SfM reconstructions, yielding diverse and verifiable supervision. We further propose Dynamic Correspondence Reinforcement Learning (DCRL), which combines Image-Level Viewpoint Progression and Point-Level Correspondence Curriculum to improve WBM training through verifiable rewards without explicit CoT supervision. Extensive experiments show that DCRL substantially improves ReasonMatch-Bench and transfers to related spatial benchmarks, while maintaining general visual understanding performance with modest gains on several benchmarks. --- ### AutoTail-BSFGM: Class-Balance-Aware Fine-Tuning for Chinese Scholarly Text Classification Authors: Anling Xiang, Yuwen Yang, Yang Shen Published: 2026-06-02 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.03576v2 Scholarly text classification supports literature organization, subject indexing, and research intelligence, but Chinese scholarly corpora often contain imbalanced and semantically adjacent disciplinary labels. We propose AutoTail-BSFGM, a class-balance-aware fine-tuning method that combines an automatically gated tail-prior adjustment, a weak Balanced Softmax auxiliary loss, and Fast Gradient Method adversarial regularization. The method changes only the training objective and procedure; inference uses the same single base-size encoder and linear classifier as the corresponding label-smoothed baseline. We evaluate the method on two CSL-based tasks: an abstract-to-discipline task with 67 labels and a title-to-category task with 13 categories. On the primary abstract task, AutoTail-BSFGM improves validation and lockbox accuracy under both Chinese RoBERTa-WWM and MacBERT-base. With MacBERT-base, validation accuracy increases by 0.83 percentage points and lockbox accuracy by 0.49 points, with a pooled paired McNemar signal on validation (p = 0.023). On the title task, the method improves validation accuracy by 0.70 points and validation balanced accuracy by 2.64 points; lockbox accuracy is approximately neutral while lockbox balanced accuracy improves by 1.22 points. The results support a bounded contribution: AutoTail-BSFGM improves class-balance-sensitive behavior and yields consistent gains for abstract-based scholarly classification, without uniformly improving every metric on every split. --- ### Few-Shot Prediction for Pulsar Noise with Long Short-Term Memory Network Authors: Qingye Tang, Dechao An, Haoran Peng, Yuqi Ouyang Published: 2026-06-02 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2606.03574v1 This work proposes a novel solution to predict pulsar timing residuals with limited data, addressing the critical challenge of data scarcity across spin-frequency subgroups of millisecond pulsars in PTA datasets. The proposed solution applies a Long Short-Term Memory (LSTM) network optimized using the model-agnostic meta-learning algorithm, enabling rapid adaptation to new frequency domain by fine-tuning the LSTM network with only a few-shot of ground truth timing residuals. Particle swarm optimization algorithm is also used for automatic hyperparameter optimization, leading to improved prediction accuracy. Our solution, evaluated on the second data release of the International Pulsar Timing Array (IPTA), demonstrates robust generalization with accurate predictions in three metrics across high-frequency test frequency domains, while requiring only 10% of the timing residuals from these domains for model fine-tuning. Furthermore, our lightweight structure only costs 16.86 MB CPU memory and 18 milliseconds for single-step residual prediction. All these characteristics make our solution highly suitable for real-world applications, where effective and real-time predictions of pulsar timing residuals are essential-particularly in resource-constrained environments with limited computational power, memory, or energy availability. --- ### When Attention Collapses: Stage-Aware Visual Token Pruning from Structure to Semantics Authors: Jiahui Wang, Kai Zhang, Mai Han, Huanghe Zhang Published: 2026-06-02 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.03569v1 Vision-Language Models (VLMs) have demonstrated remarkable capabilities but suffer from significant computational overhead during inference. While visual token pruning offers a promising solution, existing methods predominantly rely on initial attention scores. This single-metric paradigm presents a critical flaw: high attention scores inherently collapse onto semantically similar regions, thereby severely reducing feature diversity and discarding vital contextual details. To address this, we introduce Structure-to-Semantics (STS), a novel two-stage visual token pruning framework that explicitly decouples the pruning process. The first stage employs a repulsion-based sampling mechanism to maximize spatial and structural diversity. The second stage leverages instruction-aware cross-attention to precisely filter out prompt-irrelevant tokens. This two-stage synergy constitutes the core of STS, first ensuring geometric coverage and then refining the retained tokens according to semantic relevance. Extensive evaluations demonstrate that STS mitigates the redundancy caused by attention-based selection, improving both structural diversity and fine-grained task alignment of the preserved visual tokens. --- ### Learned Non-Maximum Suppression for 3D Object Detection Authors: Timo Osterburg, Stefan Schütte, Torsten Bertram Published: 2026-06-02 Categories: cs.CV, cs.AI, cs.LG, cs.RO arXiv: https://arxiv.org/abs/2606.03568v1 Post-processing is a critical stage in LiDAR-based 3D object detection, where dense and overlapping proposals must be filtered for compact and reliable perception. This work introduces two learned filtering modules that replace heuristic non-maximum suppression (NMS) by leveraging relations among detections. D2D-Rescore employs transformer-based detection-to-detection (D2D) attention, while GossipNet3D adapts the 2D GossipNet concept to 3D through localized message passing in bird's-eye view. A metric-aware matching strategy aligned with the nuScenes evaluation protocol ensures consistent training and validation behavior, improving overall detection performance. Both approaches improve mean average precision (mAP), nuScenes detection score (NDS), and true positive quality compared to CircleNMS, particularly for small and infrequent classes, while adding minimal computational overhead. These results demonstrate that learned, detection-level filtering can enhance 3D detector reliability without modifying the base network, offering a principled alternative to heuristic suppression. Code is available at https://github.com/rst-tu-dortmund/learned-3d-nms . --- ### Efficient Transformer-Based Localized Patch Sampling for Choroid Plexus Segmentation in Multiple Sclerosis Authors: Po-Jui Lu, Alessandro Cagol, Mario Ocampo-Pineda, Federico Spagnolo, Marina Mastantuono, Andreea-Alexandra Aldea, Jannis Müller, Özgür Yaldizli, Matthias Weigel, Lester Melie-Garcia, Roberta Magliozzi, Maria Pia Sormani et al. Published: 2026-06-02 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.03566v1 Background: The lateral ventricle choroid plexus (LVCP) is gaining recognition as a key imaging biomarker for multiple sclerosis (MS) related to physical disability and neuroinflammation. Yet, manual segmentation of the LVCP is highly tedious, restricting its use in broad clinical trials and longitudinal assessments. This research aims to develop a SwinUNETR-driven pipeline that leverages targeted intra- and peri-ventricular small patch sampling to automatically segment the LVCP in MS from both standalone and multi-modal MRI inputs. Methods: We retrospectively assessed 3T MRI scans across three sets of data stemming from two separate MS-dominant cohorts (Dataset 1: n=177; Dataset 2: n=177; expanded test set: n=388). Our method employed a SwinUNETR architecture trained on 32x32x32 voxel patches, benchmarking it against the 3D UXNET model. The primary metric for evaluation was the Dice Similarity Coefficient (DSC), supplemented by computational demand (GFLOPs) and the 95th percentile Hausdorff Distance (HD95). Results: On the extended test set, the SwinUNETR model secured a mean DSC of 0.868 (95% CI: 0.863-0.872) with MPRAGE and FLAIR combined, showing a statistically significant gain over UXNET (DSC: 0.858 (95% CI: 0.853-0.862), p<0.0001). When restricted to standalone FLAIR inputs, the transformer-based approach sustained a high DSC of 0.863, while the spatial localization of UXNET worsened considerably (HD95: 1.86 vs. 3.00 mm). Importantly, the proposed framework lowered computational load by 99% (91.8 vs. 22,080 GFLOPs). By integrating localized patch sampling with a SwinUNETR architecture, this methodology offers an accurate, robust, and statistically superior alternative to current leading models for LVCP segmentation. Its vast reduction in computational cost makes it ideal for widespread implementation in clinical and research environments. --- ### CR-Seg: Attention-Guided and CoT-Enhanced Coarse-to-Refined Reasoning Segmentation Authors: Yifan Cao, Xiaocui Yang, Faxian Wan, Shi Feng, Daling Wang, Yifei Zhang Published: 2026-06-02 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.03564v2 Reasoning segmentation aims to segment target objects described by complex language through joint visual-textual reasoning. Existing methods typically rely on either learned semantic tokens to bridge Multimodal Large Language Models (MLLMs) and segmentation models, suffering from difficult cross-modal alignment, or explicit spatial prompts such as bounding boxes, which may lose holistic response semantics. To address these limitations, we propose Attention-Guided and CoT-Enhanced Coarse-to-Refined Reasoning Segmentation, termed CR-Seg, a two-stage framework for coarse-to-refined reasoning segmentation. Specifically, we design an Extract Attention Maps and Points (EAP) module to extract attention maps for coarse target localization and select informative points, both of which are fed into SAM for mask refinement. To alleviate reasoning--answer inconsistency, we further introduce Global-to-Local Chain-of-Thought (GLCoT), which guides the model to reason progressively from global scene context to local target details. Extensive experiments on reasoning segmentation benchmarks demonstrate the effectiveness of CR-Seg. --- ### Analytical Evaluation of DCA Convergence Properties for Minimizing Prediction Functions of Gaussian RBF Support Vector Regression Authors: Yohei Kakimoto, Yuto Omae, Hirotaka Takahashi Published: 2026-06-02 Categories: cs.LG, math.OC, stat.ML arXiv: https://arxiv.org/abs/2606.03559v2 For nonconvex optimization problems whose objective is the prediction function of a trained Support Vector Regression (SVR) model with the Gaussian radial basis function (RBF) kernel (RBF-SVR), we present a framework that applies the difference of convex functions (DC) algorithm (DCA) by exploiting the analytical structure of the RBF kernel to construct an explicit DC decomposition. Specifically, we derive in closed form both the lower bound $μ$ of the strong convexity parameter of the DC components and the upper bound $L$ of the gradient Lipschitz constant of the subproblem. Both $μ$ and $L$ are determined solely by the post-training dual-coefficient sum $C_α$ and the RBF kernel parameter $γ$, together with the DC decomposition parameter $ρ$, and they share a common leading term $C_αρ$. Through numerical experiments on six benchmark functions, we show that $C_αρ$ is the primary single quantity characterizing both the convergence properties and the initial-point dependence of DCA, and further demonstrate that it decomposes into two independent pathways, $C \to C_α$ and $γ\to ρ$, with its primary variation governed by the SVR hyperparameters $(C, γ)$. Together, these results allow the convergence properties of DCA on RBF-SVR to be assessed in advance through the single scalar quantity $C_αρ$: approximately from $(C, γ)$ before training, and exactly in closed form after training. --- ### From Prompt to Service: An SLM-Based Agent Orchestration Gateway for AI-Driven Virtual Worlds Authors: Louis Nisiotis, Aimilios Hadjiliasi Published: 2026-06-02 Categories: cs.AI, cs.HC arXiv: https://arxiv.org/abs/2606.03557v2 As generative AI capabilities expand, AI-driven virtual worlds face a growing architectural challenge. Users interact through in-world interfaces in multimodal ways, yet their requests demand fundamentally different AI backend models and computational resources. Embedding these capabilities directly into virtual world systems reduces extensibility, complicates maintenance, and limits the ability to coordinate services distributed across edge and cloud infrastructure. This paper presents an SLM-based Agent Orchestration Gateway, a lightweight runtime coordination mechanism that decouples a virtual world client from heterogeneous AI backends through intent-driven service routing. An edge-deployed SLM classifies the semantic intent of each user prompt, a configurable service registry validates and resolves the routing decision, and the selected backend is invoked transparently, enabling new AI capabilities to be introduced in the virtual world without modifying the client application. The gateway is implemented and evaluated within the InterwovenXR virtual museum testbed. The evaluation shows that compact SLMs can serve as reliable intent routers on edge hardware, and that task-specific fine-tuning can transform sub-billion-parameter models into practical, low-latency routers. A layered configuration pairing a fine-tuned sub billion-parameter model as router with a larger SLM for conversational response generation is shown to be deployable on mid-range edge hardware and more efficient than delegating both responsibilities to a single model. The findings show that SLMs can support practical AI service orchestration in virtual worlds and the work contributes an evaluated architecture for scalable, extensible, and edge-supported AI interaction, enabling virtual agents become access points to distributed generative AI services. --- ### From Prompt to Service: An SLM-Based Agent Orchestration Gateway for AI-Driven Virtual Worlds Authors: Louis Nisiotis, Aimilios Hadjiliasi Published: 2026-06-02 Categories: cs.AI, cs.HC arXiv: https://arxiv.org/abs/2606.03557v1 As generative AI capabilities expand, AI-driven virtual worlds face a growing architectural challenge. Users interact through in-world interfaces in multimodal ways, yet their requests demand fundamentally different AI backend models and computational resources. Embedding these capabilities directly into virtual world systems reduces extensibility, complicates maintenance, and limits the ability to coordinate services distributed across edge and cloud infrastructure. This paper presents an SLM-based Agent Orchestration Gateway, a lightweight runtime coordination mechanism that decouples a virtual world client from heterogeneous AI backends through intent-driven service routing. An edge-deployed SLM classifies the semantic intent of each user prompt, a configurable service registry validates and resolves the routing decision, and the selected backend is invoked transparently, enabling new AI capabilities to be introduced in the virtual world without modifying the client application. The gateway is implemented and evaluated within the InterwovenXR virtual museum testbed. The evaluation shows that compact SLMs can serve as reliable intent routers on edge hardware, and that task-specific fine-tuning can transform sub-billion-parameter models into practical, low-latency routers. A layered configuration pairing a fine-tuned sub billion-parameter model as router with a larger SLM for conversational response generation is shown to be deployable on mid-range edge hardware and more efficient than delegating both responsibilities to a single model. The findings show that SLMs can support practical AI service orchestration in virtual worlds and the work contributes an evaluated architecture for scalable, extensible, and edge-supported AI interaction, enabling virtual agents become access points to distributed generative AI services. --- ### Constraint-Enhanced Physical Search through Correlation Matching Authors: Song-Ju Kim Published: 2026-06-02 Categories: cond-mat.stat-mech, cs.AI, nlin.AO, physics.comp-ph arXiv: https://arxiv.org/abs/2606.03554v1 Physical systems do not merely add noise to search processes; they impose constraints that generate structured correlations. We propose a principle of constraint-enhanced physical search in which temporal correlations in exploration are matched to constraint-induced spatial correlations in the update dynamics. Using a minimal tug-of-war bandit model (TOW), we show that a conservation law converts local observations into differential evidence across alternatives, while a temporally correlated drive controls the order of exploration. Search efficiency is improved not by stronger randomness or by maximal anti-correlation, but by matching the temporal correlation to the physical update scale that converts feedback into evidence. A scaling estimate identifies the update-noise-to-contrast ratio as the leading parameter that limits how strongly temporal anti-correlation can be used. The results suggest a general organizing principle for physical search: constraints and fluctuations can generate structured spatiotemporal correlations, and efficient exploration emerges when these correlations are matched to the update dynamics. --- ### A Robust Optimization Approach to Sparse Principal Component Analysis Authors: David Vävinggren, Francis Bach, André M. H. Teixeira, Dave Zachariah, Antônio H. Ribeiro Published: 2026-06-02 Categories: stat.ML, cs.LG, math.OC arXiv: https://arxiv.org/abs/2606.03553v1 While principal component analysis (PCA) is a fundamental tool for dimensionality reduction, its dense representations make it ill-suited for high-dimensional data. Existing methods address this by promoting sparsity through explicit $\ell_1$-penalties, but these are not obvious to tune due to the unsupervised nature of the task. In contrast, we propose Adversarial PCA (AdvPCA), which leverages robust optimization to achieve sparsity by optimizing the reconstruction objective against bounded, worst-case latent space perturbations. We show that this formulation admits a closed-form reduction, leading to a practical iterative algorithm that alternates between adversarial linear regression-style updates for the sparse encoder and orthogonal updates for the decoder. By theoretically characterizing the solution, we derive a data-adaptive parameterization that allows the algorithm to perform effectively out of the box. We validate these claims through numerical experiments on synthetic and real-world genomics data. --- ### How Many Trees in a Random Forest? A Revisited Approach with Plateau Search and Optuna Integration Authors: Vadim Porvatov, Andrey Dukhovny, Andrey Lange Published: 2026-06-02 Categories: cs.LG, math.PR arXiv: https://arxiv.org/abs/2606.03549v2 Hyperparameter optimization (HPO) for Random Forest faces a specific difficulty in tuning the number of trees: the predictive score typically improves monotonically with ensemble size, so standard methods such as Tree-structured Parzen Estimator (TPE) and Hyperband require a predefined search range and often drive the estimate toward its right boundary. Early-stopping strategies avoid fixing such a range, but can be sensitive to score noise and prone to premature stopping. To address this, we propose an integrated triplet-based plateau-search algorithm that removes the number of trees from the direct TPE search space and still exploits information accumulated across HPO trials. The method adaptively tracks a near-minimal sufficient ensemble size by monitoring relative changes in the out-of-bag (OOB) score across a triplet of forest sizes and shifting this triplet accordingly. This yields an automated and user-interpretable procedure based on a tolerance parameter. We also provide a theoretical analysis: we relate the proposed relative OOB-score criterion to the gap between the current and limiting scores, and derive an asymptotic variance estimate for the corresponding OOB-based absolute relative difference. Experiments show that the selected number of trees can differ substantially from the common heuristic: for most classical benchmark datasets it is smaller, whereas for some high-dimensional bioinformatics datasets, such as Arcene and Dorothea, it is larger. The source code and reproducible experiments are available at https://github.com/lange-am/rf_plateau_hpo. --- ### SAGE: A Quantitative Evaluation of Socialized Evolution in Agent Ecosystems Authors: Linyue Pan, Yaoming Zhu, Lin Qiu, Xuezhi Cao, Xunliang Cai Published: 2026-06-02 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2606.03544v1 Self-improving language agents are typically evaluated in isolation: an agent attempts a task, receives feedback, and iteratively refines its own behavior. Yet agents increasingly operate alongside peers whose strategies and outcomes are publicly visible. This raises an under-studied question: when does shared experience produce improvements that self-improvement alone cannot achieve? We introduce SAGE (Social Agent Group Evolution),an evaluation framework that compares two compute-matched conditions: SocialEvo, where agents from five distinct model families co-evolve with access to all peers' histories; and SelfEvo, where each agent receives the same number of task attempts but sees only its own past, which is conventional in self-improving agent studies. We instantiate SAGE in three arenas: open-ended ML research, long-horizon economic planning, and strategic multiplayer play, evaluated across multiple evolutionary rounds. We find that group history is not a universal amplifier: the strongest agent does not exceed its self-evolution ceiling. However, agents that plateau under self-improvement can achieve significant breakthroughs when peer experience is available. In competitive settings, counterfactual controls reveal that agents improve generally rather than developing opponent-specific strategies. Across different forms of shared history, filtered peer traces and reflective summaries often outperform raw logs, indicating that social gains depend on abstraction rather than exposure volume. These findings reveal that peer-history gains are agent-specific, arena-dependent, and contingent on the capacity to abstract transferable knowledge from public traces. --- ### Attend to Anything: Foundation Model for Unified Human Attention Modeling Authors: Wenzhuo Zhao, Ronghao Xian, Keren Fu, Qijun Zhao Published: 2026-06-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.03540v1 Existing human attention (saliency) modeling methods persist as highly fragmented across modalities, scenes, and task formulations. Consequently, even with increasing model capacity and data scale, current models predominantly remain scene-dependent and task-specific, failing to practically generalize in real-world applications. To address the fundamental limitations, we present the Attend to Anything Model (AAM), a multi-modal foundation model that unifies attention modeling across various image, video, and audio-visual tasks and scenes. AAM reformulates attention as a cognitive entailment relationship organized in a general-to-specific hierarchy, implemented through language prompts with hierarchical embeddings in hyperbolic space. Furthermore, to unify static image and dynamic video attention, we adopt a fluid-dynamics perspective, formulating video-frame attention as a diffusive temporal evolution governed by the Fokker--Planck equation. Extensive experiments on 16 benchmarks demonstrate that AAM consistently outperforms state-of-the-art methods by an average of 6\% across various scenarios, while achieving approximately a 4$\times$ speedup in video inference. Overall, these results demonstrate that AAM provides a principled foundation for future research on attention and saliency-related tasks. The dataset and code will be available at https://github.com/wz-zhao/Attend-to-Anything. --- ### Knowledge-Preserved Model Tuning in Null-Space for Robust Spatio-Temporal Video Grounding Authors: Haoxuan Chen, Xianqin Liu, Jian-Fang Hu Published: 2026-06-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.03539v1 Spatio-Temporal Video Grounding aims to localize object tubes based on textual queries. While recent methods have achieved remarkable success, they mainly focus on high-quality(HQ) inputs, neglecting the widespread presence of low-quality(LQ) videos in real-world scenarios. Although tuning methods like LoRA can adapt to degraded inputs, they inevitably disrupt pre-trained knowledge. To address this, we propose Null-Space Tuning (NST). This framework exploits the geometric property that adding vectors within the null-space of frozen weights to the layer input does not affect the output. Leveraging this, NST injects learnable residuals into input features that can be selectively invisible to the pre-trained backbone. Specifically, NST combines the Quality-Adaptive Unit and Dual-Space Reparameterization to synthesize these residuals by confining components for HQ inputs to the null-space, while directing restoration components for LQ inputs to the non-null space. As the frozen weights eliminate null-space components, we effectively rectify degraded inputs while preserving pre-trained knowledge for HQ inputs. Extensive experiments show that NST outperforms state-of-the-art methods on our Mixed-Quality benchmark. --- ### Can LLM Rerankers Predict Their Own Ranking Performance? Authors: Shiyu Ni, Keping Bi, Jiafeng Guo, Jingtong Wu, Zengxin Han, Xueqi Cheng Published: 2026-06-02 Categories: cs.IR, cs.CL arXiv: https://arxiv.org/abs/2606.03535v1 Retrieval effectiveness varies substantially across queries, making it important to estimate ranking quality before relevance judgments are available. Query performance prediction (QPP) addresses this need, but most existing methods rely on external predictors after retrieval or reranking. In this paper, we study \textit{reranker-internal QPP}: can an LLM reranker estimate the quality of the ranking it has just produced? We investigate both training-free and training-based approaches. For training-free estimation, we examine metric-specific self-consistency across sampled rankings and verbalized confidence produced directly by the reranker. Experiments on TREC Deep Learning 2019--2022 with four LLMs show that self-consistency is competitive with the state-of-the-art (SOTA) approach and better calibrated in almost all settings, while direct verbalized confidence is severely overconfident. To improve verbalized confidence, we propose two supervised methods, Verb-Num and Verb-List, which enable LLM rerankers to produce calibrated ranking-quality estimates with only a few additional output tokens. --- ### When Should the Teacher Move? Temporal Coupling and Stability in Self On-Policy Distillation Authors: Haowei Guo, Baolong Bi, Ruicheng Zhang, Bingqian Sun, Wentao Zhang Published: 2026-06-02 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.03532v1 Self on-policy distillation trains a student policy against a teacher derived from its own parameter history, yet the teacher's update schedule -- which governs the \emph{temporal coupling} between teacher and student -- has not been systematically studied as a stability variable. Through a controlled schedule sweep on Qwen3-8B, we establish that \emph{isolation periods}, defined as complete teacher freezing between updates, are the key structural property enabling stable learning, not teacher age. To characterize these underlying training dynamics, we introduce a diagnostic framework of temporal KL structure, refresh shock, and length-tail risk. This framework further uncovers \emph{state-oblivious collapse}: optimal short-horizon fixed schedules catastrophically fail under long-horizon training because a clock-driven refresh can copy a transiently drifting student into the teacher in a single, irreversible step. This failure mode is invisible under short-horizon evaluation and mechanistically distinct from EMA's chronic contamination. To address this, we propose \emph{Consolidation-Gated Teacher Refresh} (CGTR), which preserves isolation periods while gating each refresh on joint evidence of reward improvement and length-tail safety, ensuring every teacher movement responds to genuine student consolidation rather than a clock signal. With a single shared parameter set and no per-dataset retuning, CGTR achieves \textbf{zero collapse} and the best final score on all four tasks (Chemistry, Biology, Physics, ToolUse), self-regulating its refresh frequency to each task's learning dynamics. --- ### High-Precision APT Malware Attribution with Out-of-Scope Resilience Authors: Peter Williams, Adam Sobey, Erisa Karafili Published: 2026-06-02 Categories: cs.CR, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.03523v1 Early attribution of Advanced Persistent Threat (APT) activity can help defenders prioritise investigation, select countermeasures, and reduce the impact of an intrusion. Malware provides useful attribution evidence, but automated APT malware attribution remains difficult in practice. Existing approaches are typically trained and evaluated as closed-set classifiers over a limited number of known APT groups. In operational environments, however, classifiers are likely to encounter samples from groups not represented during training. Closed-set classifiers are then forced to assign such samples to known groups, producing unsupported and potentially misleading attributions. We present a high-precision APT malware attribution method based on ranked binary classifiers with explicit abstention. Rather than training a single multi-class classifier, our approach trains and tunes two binary classifiers per APT group, ranks the classifiers by validation performance, and applies them sequentially. A sample is attributed only when a classifier provides sufficient evidence; otherwise, it abstains. We evaluate the method on the APT Malware dataset and on a larger combined dataset designed to stress-test out-of-scope behaviour. On the APT Malware dataset, the method achieves higher precision than previously published results on the same dataset. In the most challenging setting, where 87% of test samples came from 60 APT groups excluded from training, the method abstained on 94% of out-of-scope samples while maintaining 92% precision and 95% selective accuracy on the samples it classified. --- ### Post-Hoc Robustness for Model-Based Reinforcement Learning Authors: Siemen Herremans, Ali Anwar, Siegfried Mercelis Published: 2026-06-02 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.03521v1 To improve the real-world applicability of reinforcement learning (RL), the field of adversarially robust RL studies how to train agents under adversarial environment perturbations. In this setting, a protagonist agent optimizes a policy under environmental perturbations from an adversary, resulting in a zero-sum Markov game. When adversarially robust RL is combined with model-based RL, the adversary can target a learned transition model instead of the training environment. Extending this idea, this work introduces post-hoc robustification of deep RL agents at inference time. By using the learned model in combination with a trained nominal policy, our approach performs a robust policy improvement step. The goal is to improve robustness without any additional training of neural networks. Specifically, we utilize model-predictive control under adversarial rollouts, which are approximated via projected gradient descent within a bounded uncertainty set. Furthermore, these offline rollouts are performed while considering and mitigating out-of-distribution issues. The proposed methodology is validated by demonstrating significant improvements in robustness when the algorithm is evaluated in perturbed Gymnasium MuJoCo environments, while considering the computational limitations of the post-hoc inference setting. --- ### Overlaying Governance: A Compositional Authorization Framework for Delegation and Scope in Agentic AI Authors: Amjad Ibrahim, Yong Li Published: 2026-06-02 Categories: cs.AI, cs.CR arXiv: https://arxiv.org/abs/2606.03518v1 As AI systems evolve from passive models into autonomous active agents capable of initiating actions, collaborating, and delegating tasks, the traditional boundaries of software systems blur. Traditional authorization and delegation frameworks, built around fixed principals, explicit requests, and static scopes, are insufficient to govern agentic systems. Agentic AI demands richer authorization semantics: agents must inherit and delegate permissions, act under time-limited authority, and coordinate through shared protocols. Existing Identity and Access Management (IAM) systems fail to fully capture this notion of agency, lacking mechanisms for recursive delegation, contextual boundaries, and dynamic scoping as executable governance primitives. Unlike access delegation standards such as OAuth 2.0, we treat delegation as a contractual term rather than merely a static token-based consent credential. This paper proposes a compositional governance framework that introduces primitives indispensable for agentic AI. We define types of delegation and their permissions and accountability implications, and we introduce a notion of resource scope attenuation to bound agentic access envelopes. These concepts are expressed as general relational definitions that can be composed into existing authorization domains (e.g., financial systems). To operationalize this composition, we define a compositional operator that overlays new agentic semantics, such as recursive delegation chains, onto existing relational policies without rewriting them. We substantiate this framework through formal proofs and empirical evaluation, showing that it provides a formal yet practical foundation for accountable authorization in agentic AI systems. --- ### Scalable On-Hardware Training of Quantum Neural Networks and Application to Clinical Data Imputation Authors: Natansh Mathur, Panagiotis Kl. Barkoutsos, Masako Yamada, Martin Roetteler, Iordanis Kerenidis Published: 2026-06-02 Categories: quant-ph, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2606.03517v1 Training quantum neural networks (QNNs) on quantum hardware is currently bottlenecked by the cost of gradient estimation: standard parameter-shift methods require a number of circuit evaluations that grows quadratically with the number of trainable parameters, making hardware-based optimisation impractical beyond small system sizes. In this work, we introduce a training framework that reduces this cost to logarithmic in the number of qubits, making gradient-based QNN optimisation feasible on near-term hardware at increasing scales. Our framework combines three co-designed ingredients: (i) a structured, subspace-preserving Butterfly circuit architecture with $O(n \log n)$ parameters and logarithmic depth; (ii) a layer-wise training strategy that confines on-hardware optimisation to one small, well-structured layer at a time; and (iii) a parallelised parameter-shift rule that exploits the commuting structure within each Butterfly layer to extract all gradients in a constant number of circuit executions. Together these reduce the number of distinct circuit evaluations per optimisation step from $O(n^2)$ to $O(\log n)$. We validate the framework on clinical data imputation using the MIMIC-III electronic health record dataset, a demanding benchmark sensitive to optimisation instability and model variance. Hybrid classical-quantum models are trained directly on IonQ Forte Enterprise trapped-ion hardware at 16 qubits without performance degradation relative to ideal or noisy simulation and via tensor-network simulation at 32 qubits, with 32-qubit inference executed on hardware. The resulting models match or exceed strong classical neural baselines in downstream patient survival prediction while exhibiting reduced variance across runs, demonstrating that the proposed framework enables practical, scalable QNN training under realistic hardware constraints. ---