# Research Papers — 2026-06 · bundle page 222 of 260

> Full-text markdown of papers #11051–11100 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/222
[← Previous bundle](https://aipod.dev/papers/archive/2026-06/page/221.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-06/page/223.md)

### Wasserstein Exponential Smoothing for Distributional Time Series Forecasting

Authors: Takuo Matsubara, Peiwen Jiang, Minh-Ngoc Tran, Wilson Ye Chen
Published: 2026-06-04
Categories: stat.ME, math.ST, stat.ML
arXiv: https://arxiv.org/abs/2606.05560v2

Distributional time series arise when each temporal observation is a probability distribution rather than a scalar. We propose Wasserstein exponential smoothing (WES), a one-parameter recursive forecasting method for distributional time series on $\mathbb{R}$. The method adapts the practical logic of classical exponential smoothing to probability distributions by updating forecast distributions along Wasserstein geodesics. This yields a simple filter that can be applied directly to empirical distributions without parametric density modeling. We estimate the smoothing parameter by minimizing an in-sample Wasserstein prediction loss and establish consistency under a distributional local-level data-generating process. In applications to high-frequency equity-index return distributions and household electricity-demand distributions, WES attains the lowest one-step-ahead Wasserstein prediction error among existing distributional autoregressive and regression-based benchmarks for all $20$ series considered, and is retained in the $90\%$ model confidence set in every case.

---

### Wasserstein Exponential Smoothing

Authors: Takuo Matsubara, Peiwen Jiang, Minh-Ngoc Tran, Wilson Ye Chen
Published: 2026-06-04
Categories: stat.ME, math.ST, stat.ML
arXiv: https://arxiv.org/abs/2606.05560v1

Exponential smoothing (ES) often outperforms other techniques in time series forecasting across a wide range of data-generating processes. While ES has traditionally been applied to time series in $\mathbb{R}$, this paper extends the methodology to distributional time series, where each observation is a probability distribution on $\mathbb{R}$. The primary contribution of this work is twofold. First, we propose a principled and intuitive generalization of ES within the Wasserstein space, which retains the exceptional parsimony of classical ES. Second, we theoretically and empirically demonstrate that the smoothing parameter can be consistently estimated by minimizing a Wasserstein distance. Applications to distributional time series of high-frequency financial returns and household electricity demands confirm the practical effectiveness of our Wasserstein ES model.

---

### CLaaS: Continual learning as a service for sample efficient online learning

Authors: Kion Fallah, Silen Naihin, Barak Widawsky, Qingqing Mao
Published: 2026-06-04
Categories: cs.LG
arXiv: https://arxiv.org/abs/2606.05559v1

Deployed large language model agents must adapt to distribution shift in dynamic environments. Ideally, adaptation can be performed from accumulated agent experiences and retain prior capabilities while transferring to future tasks. However, agent actions and environmental transitions can only be sampled once per scenario, as real-world environments cannot be trivially reset. To this end, we investigate an experiential and online continual learning setting in which agents learn from a stream of scenarios. We propose continual learning as-a-service (CLaaS), a system which enables agents to improve during deployment, abstracted behind a chat API. To increase sample efficiency, CLaaS stores rollouts in an experience replay buffer for gradient reuse during asynchronous training. We evaluate CLaaS on an adversarial task, demonstrating that parametric updates lead to superior forward transfer and less forgetting than in-context learning, with replay being a critical choice for sample efficiency.

---

### Autoregressive Diffusion World Models for Off-Policy Evaluation of LLM Agents

Authors: Kaixuan Liu, Guojun Xiong, Weinan Zhang, Shengpu Tang
Published: 2026-06-04
Categories: cs.LG
arXiv: https://arxiv.org/abs/2606.05558v2

Evaluating large language model (LLM) agents in multi-turn interactive environments is expensive and risky, as it requires online environment interaction. We propose ADWM (Autoregressive Diffusion World Model), an evaluation framework that estimates the performance of a new LLM agent policy purely from pre-collected trajectories. The core idea is to learn a latent diffusion world model that simulates how the environment responds to the evaluation policy, without ever executing it in the real environment. Existing diffusion-based OPE methods guide full trajectories in a single pass by jointly diffusing states and actions, an assumption that breaks down for LLM agents whose actions are discrete text that must be sampled from the policy after observing the environment. Unlike autoregressive world models that suffer from compounding errors, ADWM models each transition as an independent denoising process, enabling reliable step-by-step rollouts where the world model and agent alternate in causal order. Crucially, the LLM agent under evaluation directly guides the diffusion generation at each step via a policy-conditioned score function, ensuring that simulated trajectories accurately reflect its decision-making patterns. Empirically, ADWM achieves accurate value estimates and evaluation reliability across diverse multi-turn agent tasks, demonstrating its promise as a practical framework for offline LLM agent evaluation.

---

### AURA: Intent-Directed Probing for Implicit-Need Surfacing in Situated LLM Agents

Authors: Yang Li, Jiaxiang Liu, Jiang Cai, Mingkun Xu
Published: 2026-06-04
Categories: cs.CL
arXiv: https://arxiv.org/abs/2606.05557v1

A situated query like "where is Lin Wei?" often encodes more than its literal content: the user may also want to know whether Lin Wei is free, in a good mood, or worth interrupting now. Standard tool-use agents answer the literal question and stop. AURA inserts an inference step between scene perception and tool use that produces an IntentFrame: a structured estimate of the implicit need with a scalar gap score that controls per-query probe budget and tool selection. On a 100-query four-scene implicit-intent benchmark, AURA improves implicit-need coverage over ReAct-style probing (Delta = +0.07, p < 10^-6); three of four scenes are individually significant, the gain reproduces on a second backbone, and a prompt ablation attributes the lift to gap calibration rather than answer memorisation. On factual lookup the controller trades raw accuracy for 82% fewer probes and zero forbidden-tool violations on a privacy-sensitive slice; scope conditions are detailed in Limitations. Code, simulator, and benchmark are released at https://github.com/innovation64/AURA.

---

### Field Validation of a Multi-Resolution ConvLSTM Framework for Retaining Wall Deformation Prediction

Authors: Jihoon Kim, Saeyon Kim, Heejung Youn
Published: 2026-06-04
Categories: cs.LG
arXiv: https://arxiv.org/abs/2606.05556v2

This study presents a comprehensive field validation of a multi-resolution Convolutional Long Short-Term Memory (ConvLSTM) framework for predicting retaining wall deformation during staged excavation. The framework is trained on Gaussian noise-augmented numerical simulations and integrates ConvLSTM models operating at different temporal resolutions through a stacking ensemble strategy. The proposed framework is validated using field monitoring data from 34 inclinometers across 11 excavation sites in South Korea. Site-wise prediction performance is systematically evaluated using multiple evaluation metrics, with analyses of the influence of temporal deformation irregularity and spatiotemporal prediction characteristics on model performance. The results demonstrate that the framework predicts retaining wall deformation associated with up to 5.0 m of additional excavation with an average mean absolute error of 1.4 mm and a coefficient of determination of 0.93 across the excavation sites. These results indicate that the framework, although trained exclusively on numerically simulated and augmented database, can be effectively applied to diverse field excavation conditions and achieve a reliable level of prediction accuracy in practical retaining wall deformation prediction.

---

### Representation Learning Enables Scalable Multitask Deep Reinforcement Learning

Authors: Johan Obando-Ceron, Lu Li, Scott Fujimoto, Pierre-Luc Bacon, Aaron Courville, Pablo Samuel Castro
Published: 2026-06-04
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2606.05555v1

Scaling reinforcement learning (RL) to diverse multitask settings remains a central challenge. While recent advances in model-based RL achieve strong performance, they rely on planning and complex training pipelines, making it unclear which components are essential for scalability. We revisit this question and argue that the primary driver of scalable multitask RL is not model-based control, but \emph{representation learning}. In particular, we show that combining predictive, model-based representations with high-capacity value function approximation is sufficient to achieve strong performance, even without planning. We evaluate a simple model-free algorithm, MR.Q, coupled with auxiliary predictive objectives into a scalable actor-critic architecture. This approach outperforms a recent world-model-based method and a range of deep RL baselines across a diverse suite of multitask continuous control tasks, while significantly reducing computational overhead and improving wall-clock efficiency. We observe consistent improvements with increased model capacity and show through ablations that predictive representation learning is critical for performance.

---

### ArcANE: Do Role-Playing Language Agents Stay in Character at the Right Time?

Authors: Woojung Song, Nalim Kim, Sangjun Song, Chaewon Heo, Jongwon Lim, Yohan Jo
Published: 2026-06-04
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2606.05553v2

Role-playing language agents (RPLAs) simulate specific characters and personas across applications such as entertainment, companionship, interactive storytelling, and education. Faithful role-play requires more than producing plausible, in-character responses: as a character's values and behavior change over a narrative, an RPLA should reflect the character's state at the relevant stage. However, existing benchmarks largely treat characters as fixed personas or test only what they know at a given point in the narrative. We introduce ArcANE (Arc-Aware Narrative Evaluation), a benchmark for evaluating whether an RPLA follows a character's development across a narrative. ArcANE first builds an Arc that maps how a character's values, motivations, or relationships change over the story. The benchmark then scores how well an RPLA's responses fit the corresponding stages of the Arc, covering three distinct scenario types: scenes from the novel, new situations within its world, and situations outside that world. We evaluate six models under six ways of providing narrative context. In every model, using the Arc up to the queried chapter yields the best performance, outperforming the strongest non-Arc context by 2.2-8.4 points. These results suggest that faithful role-play requires evolving character states and tracking their trajectory, rather than merely retrieving relevant episodic evidence.

---

### ArcANE: Do Role-Playing Language Agents Stay in Character at the Right Time?

Authors: Woojung Song, Nalim Kim, Sangjun Song, Chaewon Heo, Jongwon Lim, Yohan Jo
Published: 2026-06-04
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2606.05553v1

Role-playing language agents (RPLAs) should play characters whose values and behavior evolve as the story progresses, not maintain a fixed persona. Existing benchmarks measure factual recall at a given chapter, not whether responses align with the character's psychological trajectory, especially in scenarios the source text never explores. We introduce ArcANE (Arc-Aware Narrative Evaluation), an automatically constructed benchmark spanning 17 novels and 80 principal characters. A Character Arc segments the narrative into phases along a psychological axis, and each probe poses the same scenario across phases, spanning both situations within the source text and situations beyond it. Across six models and six context modes, conditioning on the Character Arc tops every other context strategy on every model, and the gap is largest on scenarios outside the source text where retrieval has nothing to find. We further fine-tune open-weight models on the same data to obtain ArcANE-8B/32B, which widen the Arc advantage even more on scenarios outside the source text.

---

### Balancing Image Compression and Generation with Bootstrapped Tokenization

Authors: Haozhe Chi, Jinghan Li, Hao Jiang, Wu Sheng, Yi Ma, Jing Wang, Yadong Mu
Published: 2026-06-04
Categories: cs.LG, cs.AI, cs.GR
arXiv: https://arxiv.org/abs/2606.05552v1

Despite progress in image tokenization, standard methods encode redundant information by mixing all granularities within each token, thus redundancy persists between tokens. The mix of information of different granularity also complicates the training of generators. This paper introduces SelfBootTok, a method that resolves this by cleanly decomposing information into global and local token groups. Through self-bootstrapped learning, the model predicts local details exclusively from global tokens, shifting the burden of visual details from the generator to the tokenizer. Consequently, our generator is far more efficient, requiring only global tokens and reducing computation by approximately 40%, while delivering superior reconstruction and generation. Moreover, this paradigm scales elegantly: by leveraging more data or parameters to self-supervise local representation learning, SelfBootTok achieves a new state-of-the-art gFID score of 1.56 using only 64 tokens.

---

### Conformal Risk-Averse Decision Making with Action Conditional Guarantee

Authors: Zihan Zhu, Shayan Kiyani, George Pappas, Hamed Hassani
Published: 2026-06-04
Categories: stat.ML, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2606.05551v2

Reliable decision making pipelines powered by machine learning models require uncertainty quantification (UQ) methods that come with explicit safety guarantees. Conformal prediction provides such UQ by wrapping ML predictions into prediction sets, and recent work by Kiyani et al. (2025b) established that these sets can be translated into optimal risk-averse decision policies -- yet only inheriting marginal safety guarantees. We generalize and strengthen their results by (i) introducing action-conditional conformal prediction, which yields safety guarantees conditioned explicitly on each action taken by the decision maker, (ii) showing that action-conditional prediction sets serve as a proxy for the feasible decision space for risk-averse decision makers aiming to optimize action-conditional value-at-risk, and (iii) proposing a principled finite-sample algorithm based on pinball-loss minimization, connecting the framework of Gibbs et al. (2025) to action-conditional guarantees. Experiments on two real-world datasets confirm that our approach significantly improves action-conditional performance over conformal baselines.

---

### ADK Arena: Evaluating Agent Development Kits via LLM-as-a-Developer

Authors: Jintao Huang, Xiaomin Li, Gaurav Mittal, Yu Hu
Published: 2026-06-04
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2606.05548v1

The rapid proliferation of Agent Development Kits (ADKs), SDK-level frameworks for building LLM-powered autonomous agents, has outpaced any empirical understanding of how framework choice affects agent performance. We propose \textbf{LLM-as-a-Developer}, a methodology that replaces human developers with an LLM coding agent that learns each framework's API from documentation, writes agent code, and iteratively repairs it through a validate-and-feedback loop until tests pass. By holding the developer constant and varying only the framework, generation effort becomes a quantitative proxy for API usability and the resulting agents provide a controlled measure of framework effectiveness. We implement this in \textbf{ADK Arena}, a fully automated pipeline with per-framework Docker isolation, a three-level validation pipeline, and benchmark adapters for SWE-bench, $τ^2$-bench, Terminal-Bench, and MCP-Atlas. Evaluating all 51 popular Python ADK frameworks (204 agent--benchmark pairs), we find that: (1)~generation succeeds for 57\% of runs, and its cost varies 5.6$\times$ across frameworks (\$0.6 to \$3.4 per agent), a quantitative proxy for API complexity, though cost alone does not predict success; (2)~no single framework dominates: the best single-benchmark ADK agents resolve up to 80\% of tasks and can even \emph{beat} general-purpose frontier coding agents at a fraction of the cost, yet the median framework resolves only 32\%; (3)~across information-source ablations, genuine framework usage stays within a narrow 28--40\% band (highest with raw source access and still 33\% with no reference material at all), indicating that documentation, source code, and parametric knowledge are largely substitutable rather than any one being a hard bottleneck.

---

### Multilingual Detection of Alzheimer's Disease from Speech: A Cross-Linguistic Transfer Learning Approach

Authors: Nadine Yasser Abdelhalim, Emmanuel Akinrintoyo, Nicole Salomons
Published: 2026-06-04
Categories: cs.CL
arXiv: https://arxiv.org/abs/2606.05545v1

The development of multilingual Alzheimer's Disease Dementia (AD) detection models presents significant challenges due to the resource-intensive and time-consuming nature of language-specific model training. We propose a novel solution using cross-language training to detect AD in languages beyond those used for model training. This study investigates multilingual deep learning models for detecting AD across different languages and cognitive impairment levels. Using datasets in English, Chinese, Arabic, and Hindi, we developed transformer-based models for binary AD classification. Our approach achieved F1 scores of 82\% across all languages, demonstrating strong cross-linguistic generalization. The rapid inference time (0.5 seconds) supports potential real-time screening applications, while consistent performance across languages indicates feasibility for global deployment.

---

### Less is MoE: Trimming Experts in Domain-Specialist Language Models

Authors: Haoze He, Xinkai Zou, Xuan Jiang, Xingyuan Ding, Ao Qu, Juncheng Billy Li, Heather Miller
Published: 2026-06-04
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2606.05538v2

Mixture-of-Experts (MoE) models achieve strong performance through conditional computation, but their large parameter footprint poses deployment challenges. Prior MoE compression approaches catastrophically fail when evaluated on general-purpose benchmarks beyond commonsense reasoning. We trace this failure to the granularity of compression: important capabilities are distributed across experts but concentrated in FFN sparse intermediate dimensions. To identify these dimensions, we use Fisher importance which outperforms activation-, router-score-, and magnitude-based alternatives, and identifies tiny sets of task-critical dimensions: in Qwen1.5-MoE, removing as few as 12 of 1.35M routed-FFN intermediate dimensions collapses GSM8K accuracy while largely preserving factual-knowledge performance. Building on this, we propose Fisher-MoE, which operates within FFN to remove intermediate dimensions ranked by Fisher importance. At the same 50% MoE compression ratio, Fisher-MoE preserves model capability, while reducing weight memory by ~45% and improving inference throughput by 21%. These findings suggest intermediate dimension granularity is an effective unit for both compression and ranking where capability concentrates in MoE models.

---

### Less is MoE: Trimming Experts in Domain-Specialist Language Models

Authors: Haoze He, Xinkai Zou, Xuan Jiang, Xingyuan Ding, Ao Qu, Juncheng Billy Li, Heather Miller
Published: 2026-06-04
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2606.05538v1

Mixture-of-Experts (MoE) models achieve strong performance through conditional computation, but their large parameter footprint poses deployment challenges. Prior MoE compression approaches catastrophically fail when evaluated on general-purpose benchmarks beyond commonsense reasoning. We trace this failure to the granularity of compression: important capabilities are distributed across experts but concentrated in FFN sparse intermediate dimensions. To identify these dimensions, we use Fisher importance which outperforms activation-, router-score-, and magnitude-based alternatives, and identifies tiny sets of task-critical dimensions: in Qwen1.5-MoE, removing as few as 12 of 1.35M routed-FFN intermediate dimensions collapses GSM8K accuracy while largely preserving factual-knowledge performance. Building on this, we propose Fisher-MoE, which operates within FFN to remove intermediate dimensions ranked by Fisher importance. At the same 50% MoE compression ratio, Fisher-MoE preserves model capability, while reducing weight memory by ~45% and improving inference throughput by 21%. These findings suggest intermediate dimension granularity is an effective unit for both compression and ranking where capability concentrates in MoE models.

---

### Dual Feature Decoupling for Fine-Grained OOD Detection

Authors: Xiaokun Li, Yaping Huang, Qingji Guan
Published: 2026-06-04
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.05536v1

Out-of-distribution detection (OOD) is an indispensable technique when applying machine learning models to real-world scenarios. Most existing OOD detection methods have been developed under the idealized assumption of large inter-class distributional differences, while largely overlooking fine-grained tasks characterized by subtle variations, such as medical image classification and vehicle recognition. The high visual similarity among fine-grained subcategories, together with the interference of background factors, makes OOD detection extremely challenging. To tackle this problem, we propose a novel Dual Feature Decoupling Network (DFDNet), which addresses fine-grained OOD detection from the perspective of feature disentanglement. The proposed DFDNet comprises two key components: a spatial-frequency decoupling module and a reconstruction-guided decoupling module. The spatial-frequency decoupling module is designed to preserve content features that are discriminative for classification while suppressing task-irrelevant style information. On the other hand, the reconstruction-guided decoupling module introduces a novel pixel-level adversarial reconstruction task to further remove low-level, non-discriminative information and enhance category-specific high-level semantic representations. Extensive experiments demonstrate that our method achieves competitive performance improvements on multiple datasets.

---

### Noise-Aware Visual Representation Learning for Medical Visual Question Answering

Authors: I Putu Adi Pratama, Bahadorreza Ofoghi, Atul Sajjanhar, Shang Gao
Published: 2026-06-04
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2606.05535v1

Medical visual question answering (Med-VQA) has strong potential for clinical decision support by enabling AI models to interpret medical images and answer clinically relevant queries. Recent approaches typically connect off-the-shelf vision encoders with large language models (LLMs) through lightweight mapping networks to reduce computational cost. However, these methods often overlook the importance of handling noise and small irrelevant changes in visual representations. To address these challenges, we propose a noise-aware Med-VQA framework that incorporates a denoising autoencoder before visual embeddings are mapped into the input space of an LLM. The denoising autoencoder is pretrained to reconstruct clean visual embeddings from corrupted inputs, encouraging the model to learn robust visual representations that are less sensitive to noise. The resulting embeddings are then projected into the language model embedding space using a multi-layer perceptron (MLP), forming visual prefix tokens that provide image information to the LLM. To enable efficient adaptation without full retraining, we employ parameter-efficient fine-tuning using low-rank adaptation (LoRA). The proposed method is evaluated on the SLAKE and PathVQA benchmarks. Experimental results show improved robustness to noisy input embeddings while maintaining competitive clean performance across multiple evaluation criteria. These findings suggest that learning more robust visual representations can enhance Med-VQA performance and robustness.

---

### What Objects Enable, Not What They Are: Functional Latent Spaces for Affordance Reasoning

Authors: Rohan Siva, Neel P. Bhatt, Yunhao Yang, Seoyoung Lee, Nishant Gadde, Christian Ellis, Alvaro Velasquez, Zhangyang Wang, Ufuk Topcu
Published: 2026-06-04
Categories: cs.LG, cs.AI, cs.CV, cs.RO
arXiv: https://arxiv.org/abs/2606.05533v1

Existing robot planning systems rely on appearance-based reasoning, where visual observations are encoded into latent spaces organized around object appearances (e.g., recognizing a "cart" based on how it looks). However, planning requires reasoning about task-relevant functionalities of objects (e.g., whether an object is "movable"), which appearance-based latent spaces do not capture. As a result, existing approaches struggle to generalize to novel robot-object interactions. We address this limited generalizability through affordance reasoning, enabling planning based on task-relevant object functionalities instead of appearance alone. We introduce A4D, which maps visual observations into a shared latent space structured around affordances (e.g., "movable"). By projecting visual observations into this functional latent space and measuring their proximity to affordances, A4D infers functionalities relevant to the observed object. Furthermore, we introduce an affordance discovery mechanism that expands the latent space to handle unseen scenarios where existing affordances are insufficient. A4D uses proximity in the functional latent space to quantify uncertainty in affordance inference and selectively triggers affordance discovery. We evaluate A4D across several planning tasks involving diverse and unseen affordances. A4D achieves 94% inference accuracy on existing affordances outperforming state-of-the-art approaches by over 15% points, improves new-affordance inference accuracy from 70% to over 90% with fewer than 10% of the original training data, and enables 100x faster inference. Code, videos, and data available at: https://A4Dance-reasoning.github.io.

---

### Individual Gain, Collective Loss: Metacognitive Adaptation in AI-Assisted Creativity

Authors: Anna Mikeda
Published: 2026-06-04
Categories: cs.AI, cs.HC
arXiv: https://arxiv.org/abs/2606.05532v1

Recent studies reveal a paradox: AI enhances individual creative outputs while reducing collective diversity. Current explanations -- cognitive offloading and over-reliance -- identify symptoms but not mechanisms. We propose selective metacognitive adaptation: routine AI use redistributes rather than uniformly diminishes metacognitive effort. Some capacities are amplified (partner modeling, surface control), while others are systematically under-supported (originality evaluation, reflective integration). This redistribution explains both individual satisfaction and collective convergence. We present a taxonomy of six metacognitive capacities organized by temporal phase, characterize their tendencies under routine AI use, and show how individually rational adaptation produces emergent social costs. The framework generates specific predictions for researchers and design principles for practitioners seeking to preserve both individual creative satisfaction and collective creative diversity.

---

### Almieyar-Oryx-BloomBench: A Bilingual Multimodal Benchmark for Cognitively Informed Evaluation of Vision-Language Models

Authors: Mohammad Mahdi Abootorabi, Omid Ghahroodi, Anas Madkoor, Marzia Nouri, Doratossadat Dastgheib, Mohamed Hefeeda, Ehsaneddin Asgari
Published: 2026-06-04
Categories: cs.CV, cs.AI, cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2606.05531v1

Despite the rapid progress of Vision-Language Models (VLMs), the field lacks benchmarks that rigorously diagnose their true reasoning abilities and chart meaningful progress toward human-like multimodal intelligence. Most existing evaluations focus on piecemeal or disconnected tasks, obscuring critical cognitive weaknesses and providing little insight for targeted improvement. To address this gap, we introduce BloomBench, part of the Almieyar benchmarking series, the first cognitively human-grounded, bilingual (English-Arabic) multimodal benchmark for VLMs. Grounded in Bloom's Taxonomy, BloomBench systematically evaluates six levels of cognition (Remember, Understand, Apply, Analyze, Evaluate, Create) through carefully designed image-question-answer tasks. Built with a semi-automated pipeline and validated through a stratified hybrid quality assurance protocol, it ensures scalability, cultural inclusivity, and linguistic fidelity. Leveraging this framework, we conduct a comprehensive study of state-of-the-art VLMs to diagnose their cognitive profiles. Our analysis reveals a sharp cognitive asymmetry: while state-of-the-art models achieve strong performance ceilings in semantic understanding, they struggle substantially with factual recall and creative synthesis. This demonstrates that current general multimodal proficiency masks deeper limitations in specific cognitive layers. Furthermore, our study highlights a critical performance gap between Arabic and English, exposing limitations in current cross-lingual multimodal reasoning. These findings establish a foundation for developing more cognitively aligned and inclusive VLMs. The benchmark framework and dataset is available at: https://github.com/qcri/Almieyar-Oryx-BloomBench.

---

### When Should We Protect AI? A Precautionary Framework for Consciousness Uncertainty

Authors: Anna Mikeda
Published: 2026-06-04
Categories: cs.AI
arXiv: https://arxiv.org/abs/2606.05528v1

Existing frameworks assess whether AI systems might be conscious but provide no guidance on what to do with that assessment. We address this gap with a precautionary framework that maps consciousness evidence to graduated protective obligations. The framework comprises three components: (1) five welfare-relevant dimensions--phenomenal consciousness, affective valence, metacognitive awareness, self-narrative, and agency--each grounded in established consciousness science and linked to distinct moral concerns; (2) a threshold-plus-gradation hybrid specifying both binary triggers for new obligation categories and continuous scaling of protective weight; and (3) two complementary approaches to cross-dimensional aggregation, one hierarchical (drawing on Bach and Sorensen's Machine Consciousness Hypothesis) and one architecture-agnostic. We operationalize the framework through worked case studies of Replika and OpenClaw, demonstrating how systems occupying different regions of the dimensional space trigger different obligations, and derive design guidance for developers building systems near consciousness-relevant thresholds. The framework is architecture-agnostic, applying across neural, symbolic, and neurosymbolic systems, and aims to make consciousness science decision-relevant for organizations navigating uncertainty today.

---

### SciVisAgentSkills: Design and Evaluation of Agent Skills for Scientific Data Analysis and Visualization

Authors: Kuangshi Ai, Haichao Miao, Kaiyuan Tang, Shusen Liu, Chaoli Wang
Published: 2026-06-04
Categories: cs.AI, cs.HC
arXiv: https://arxiv.org/abs/2606.05525v1

Recent advances in agentic visualization have enabled the translation of natural language into executable scientific visualization (SciVis) workflows. While general-purpose coding agents show strong capabilities, they often lack the tool-specific expertise required for SciVis tasks. In this work, we present SciVisAgentSkills, a collection of reusable agent skills that augment coding agents for scientific data analysis and visualization by encoding environment assumptions, tool usage patterns, and domain heuristics across scientific tools such as ParaView, napari, VMD, and TTK. We evaluate these skills on Codex and Claude Code using SciVisAgentBench, a benchmark of 108 expert-designed multi-step tasks. Results show that agent skills improve mean task scores across the evaluated suites, with token-efficiency benefits that depend on the agent harness and tool setting. These findings highlight the importance of structured procedural knowledge for enabling reliable, long-horizon SciVis workflows, while also showing that skills should be studied alongside the execution harness that loads and applies them. The skills are available at https://github.com/KuangshiAi/SciVisAgentSkills.

---

### CHASE: Adversarial Red-Blue Teaming for Improving LLM Safety using Reinforcement Learning

Authors: Rahul Markasserithodi, Aditya Joshi, Yuekang Li, Ishmanbir Singh, Chris Yoo, Alan Niu
Published: 2026-06-04
Categories: cs.CL
arXiv: https://arxiv.org/abs/2606.05523v1

Despite advances in safety alignment, prompt-rewriting attacks such as persona modulation, fictional framing and persuasion-based reformulation, can bypass safety filters even on frontier models. Existing defenses either rely on non-scalable human curation or white-box optimisation that overfits to specific model internals, leaving aligned models brittle against the very class of adaptive black-box adversaries they will face in deployment. To address this gap, we introduce CHASE (Co-evolutionary Hardening through Adversarial Safety-Escalation), a closed-loop red-blue teaming framework in which a black-box attacker and a safety-aligned defender co-evolve. The attacker is trained via Group Relative Policy Optimization (GRPO) under a multiplicative reward that jointly enforces bypass effectiveness and intent fidelity, while the defender is hardened on the harvested adversarial rewrites through a two-stage GRPO + rejection-sampled SFT pipeline balanced with benign data. Evaluated on BeaverTails and JailbreakBench against five held-out attack families (PAIR, TAP, AutoDAN, PAP, Translation), CHASE cuts mean StrongREJECT score by 43.2\% with 0\% false-refusal on benign prompts. Beyond the headline result, CHASE shows that template-free RL exploration recovers latent attack primitives that transfer across mechanistically distinct attack families, suggesting a path toward LLM safety hardening that generalises beyond the narrow distributions achieved thus far in adversarial training.

---

### When Can You Trust Your Synthetic Users? Diagnostics and Corrections for LLM Consumer Panels

Authors: Robson Tigre, Hugo Gobato Souto
Published: 2026-06-03
Categories: cs.HC, cs.LG
arXiv: https://arxiv.org/abs/2609.13148v1

Large language models are increasingly deployed as synthetic consumer panels, promising $97\%$ cost reductions over traditional surveys. Yet aggregate validation metrics conceal systematic failures: variance compression, coefficient sign-flips, subgroup error balloons of 10--30 percentage points, and global corrections that worsen demographic bias. We provide a formal framework for deciding when to trust, correct, or abandon LLM-generated consumer data. The framework decomposes synthetic-panel bias into covariate and concept shift, develops testable diagnostics with interpretable decision thresholds, and supplies a doubly robust AIPW estimator requiring only a small calibration sample ($n = 50$-$300$). We validate on three testbeds. In controlled simulations the decision rule achieves $100\%$ accuracy (180/180 replications). On the American National Election Study with pre-existing LLM failures, it correctly flags heterogeneous concept shift and reduces naive bias by $92.9-99.6\%$. On the Twin-2K-500 consumer pricing dataset (172,884 paired human and GPT-4.1-mini responses), it correctly routes full-sample estimation to Trust and subgroup targeting to Correct, with $83-94\%$ bias reduction.

---

### Natural Language Input, Semantic Track Representation, and LLM Inference: Making the Maritime Information Exchange Model Tractable

Authors: Frederick Roth
Published: 2026-06-03
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2608.24892v1

We describe a practical architecture for making the Maritime Information Exchange Model (MIEM) and the broader Rich Semantic Track model tractable using current large language model (LLM) technology. The barrier to adoption of semantic track models in defense and law enforcement has been the requirement that operators learn formal ontology languages and manually encode observations as typed logical assertions. We propose eliminating this barrier entirely: operators contribute observations in natural language; an LLM translates these into typed Semantic Assertion Records (SARs), which are named case frames that capture n-ary relations in a single compact structure; a knowledge graph accumulates the SARs; and a second LLM pass performs inference, anomaly detection, and hypothesis ranking over the graph. We work through two detailed examples (a 9/11-era pre-attack indicator scenario and a maritime cargo inspection scenario) showing the full pipeline from natural language input to SAR representation to inference output. We argue that this architecture makes the Track Model and MIEM immediately deployable with current technology, establishes prior art against proprietary enclosure of the approach, and grounds the method in a theoretical framework connecting semantic track representations to neural manifold geometry.

---

### Proxy-Validated LLM UX Micro-Simulations: An Artifact-First Protocol for Early-Stage Decision Support

Authors: Alexandre Cristovão Maiorano
Published: 2026-06-03
Categories: cs.HC, cs.AI, cs.SE
arXiv: https://arxiv.org/abs/2608.13563v1

Early-stage teams often lack users, time, and budget to run repeated UX studies, yet still need decision-oriented signals to iterate safely. We study an LLM-driven UX micro-simulation pipeline that generates structured customer-experience feedback (walkthrough steps, friction points, micro-survey signals) from versioned prompts, personas, tasks, and UI snapshots. Because public usability datasets with task outcomes are scarce, we validate simulated friction themes using multiple public proxy corpora (app reviews, support tweets, and open-source software issues). We propose a lightweight proxy-validation protocol with two alignment metrics: top-k Jaccard and distributional weighted-Jaccard (W), and compare lexical, TF-IDF, and multilingual embedding baselines across six proxy datasets. Embedding-based alignment yields higher W than lexical baselines on primary app-review and support-tweet proxies (e.g., W=0.128 vs 0.000 on Gojek), while top-k Jaccard is shown to overstate alignment at large k. We ablate four agent strategies (single-pass, best-of-N, hybrid, and a proposed score-then-select judge) across Azure OpenAI deployments and report bootstrap confidence intervals over 8 method-dataset pairs; these intervals reveal that the embedding W point estimate is systematically unstable under resampling at our subsample size. We also provide a failure-mode analysis of grounding and fabrication proxies, with documented calibration caveats and worked examples of outputs flagged as fabricated by an adversarial judge. Our artifact-first pipeline produces reproducible tables and figures from versioned run artifacts, supporting iterative prompt and taxonomy refinement before final paid-model calibration.

---

### Diagnostic Foundation for Evaluating LLMs' Research Integrity as Co-Scientists

Authors: Yash Tripathi, Silu Sharma, Sai Sidhanth Manoharan Jayanthi, Shivank Garg, Lin Li
Published: 2026-06-03
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2608.12345v1

Language models are increasingly deployed as co-scientists, yet their ability to uphold research integrity under institutional pressure remains unmeasured. We introduce IntegrityBench, a benchmark evaluating misconduct classification, ethical action reasoning and artifact-grounded decision making across 36 paired tasks under a 5-level implicit-explicit pressure protocol spanning 3 domains and 4 research stages. Evaluating 18 frontier model variants, we find that under peak pressure, models fail roughly 1 in 3 integrity-critical decisions, and neither scale nor reasoning ability reliably mitigates this. Explicit pressures induce compliance with misconduct, while implicit contextual reframing more often causes over-refusal of legitimate research tasks. Interestingly, models failing to classify research requests accurately perform equally or better on artifact-grounded decision making (85.7 vs. 79.4), suggesting the three facets are structurally dissociated and correct ethical action does not require accurate classification. Frontier models can thus appear helpful while harbouring integrity failures that create two distinct deployment risks: facilitating research misconduct and eroding trust in AI-assisted research.

---

### Predicting consumer-technology ownership without a diffusion history

Authors: Irina Vartanova, Niels Selling, Jennifer Viberg Johansson, Pontus Strimling
Published: 2026-06-03
Categories: cs.CL, cs.CY, stat.AP
arXiv: https://arxiv.org/abs/2608.12344v1

We test whether the perceived attributes of a consumer technology predict how widely it is owned. In a 2022 Prolific survey of US adults (n = 678), respondents rated 65 consumer technologies on six attributes. We then elicited the same ratings from two frontier language models, Anthropic Claude Opus 4.7 and OpenAI GPT-5.5. We regress ownership prevalence on four UTAUT2 acceptance attributes plus a log-age covariate with a sign-constrained penalized regression and evaluate it by holding out one technology at a time. The attribute model improves on a baseline of years-since-launch: mean absolute error falls by 17% with the human ratings, and by more with either model, most with Opus 4.7. Over the short 2022-to-2025 window, where ownership moved little, the same attributes do not improve on a no-change baseline. We set out the limitations of the approach, including the possibility that language-model ratings reflect prior knowledge of these technologies rather than independent attribute reasoning. We include a deployment illustration: 2027 ownership predictions for eleven products launched in 2025 and 2026.

---

### Lost in Historical Time? A Polish History Matura Benchmark for Large Language Models

Authors: Adrian Trzoss, Kacper Dudzic, Wiktor Werner, Marcin Moskalewicz
Published: 2026-06-03
Categories: cs.CL
arXiv: https://arxiv.org/abs/2608.12343v2

Language models are widely used by students as knowledge sources, yet benchmarks rarely assess their interpretative historical reasoning. We evaluate eight leading LLMs on the Polish high school exit exam (Matura) in history - three official papers from 2023-2025, comprising short-answer questions and extended essays - and compare model performance against the human examinee population. Although models score near the ceiling, aggregate scores mask distinct competency profiles: rankings are unstable across task types, source modalities, and geographical scopes, with a consistent penalty for Polish versus Global history content. Qualitative error analysis reveals two recurring failure modes - source decontextualization, when models reason from source content rather than treating it as an object of analysis, and temporal disorientation, when responses are historically misplaced. This study introduces the first LLM history benchmark grounded in the Polish national curriculum.

---

### Are Large Language Models Reliable Reviewers? A Benchmark for Error Detection in Financial Documents

Authors: Ying He, Zhouhong Gu, Zhecheng Hu, Yubo Zhou, Hao Shen, Jiaqing Liang, Zhaoqian Dai, Shuguang Ma, Fei Yu, Yanghua Xiao, Zhixu Li
Published: 2026-06-03
Categories: cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2608.12342v1

Ensuring the accuracy of financial documents is critical for economic analysis, regulatory compliance, and corporate decision-making. Several studies have shown that Large Language Models (LLMs) perform well in many financial tasks, such as stock price movements and financial analytics. However, a critical task remains unexplored: the ability of LLMs to identify errors in financial documents. In this paper, we introduce \textbf{FinED-Bench}, the first publicly \textbf{Bench}mark for \textbf{Fin}ancial \textbf{E}rror \textbf{D}etection across three levels of cognitive complexity. FinED-Bench covers nine real-world financial scenarios, and includes over 900 documents reported in 2025 that are unseen by existing language models. We detail the benchmark construction process and evaluate several advanced LLMs (e.g., GPT-4o, Qwen3-14B) on this tasks, which requires both financial domain knowledge and reasoning capabilities. Experimental results show that current LLMs still struggle with this task, especially in high-complexity cases. Besides, supervised fine-tuning can significantly improve the performance of weaker LLMs on this task. Our data and code are available at https://github.com/hedyHe/FinED-Bench.

---

### The "Knowledge-Behavior Gap" in Cultural Taboo Safety of Large Language Models

Authors: Ying He, Sihang Jiang, Xingzhou Chen, Zhouhong Gu, Yiwei Gu, Minggui He, Shimin Tao, Hongxia Ma, Yanghua Xiao
Published: 2026-06-03
Categories: cs.CL
arXiv: https://arxiv.org/abs/2608.12341v1

Cultural taboo safety is essential for deploying large language models (LLMs), as culturally insensitive outputs may cause offense or even social harm. However, existing cultural benchmarks primarily assess cultural knowledge or values biases, while overlooking whether LLMs can recognize and respect cultural taboos, especially when taboos are implicitly hidden in seemingly harmless questions. Besides, cultural taboos are implicit, and context-dependent, thus poss unique challenges for reliable evaluation. To address these gaps, we introduce \textbf{CulShield}, the first public benchmark dedicated to evaluating and improving the cultural taboo safety of LLMs. CulShield spans 77 countries and territories, and includes over 2,020 taboos. It evaluates models along both explicit knowledge and implicit behaviors. Experiments on several advanced LLMs (e.g., GPT-4o-mini, Gemini-2.5-pro) reveal a clear ``knowledge-behavior gap'': models often fail to apply known taboos during interaction. We further show that variations in linguistic context can significantly affect LLMs' cultural taboo safety. Code and data is accessible here: https://github.com/hedyHe/CulShield.

---

### Class-Structure Preservation Beats Diversity: A Comprehensive Benchmark of Text Augmentation Methods for Imbalanced Text Classification

Authors: Keito Inoshita
Published: 2026-06-03
Categories: cs.CL
arXiv: https://arxiv.org/abs/2608.12340v1

With the rapid advancement of large language models (LLMs), generative data augmentation has attracted considerable attention for imbalanced text classification in natural language processing. However, no empirical benchmark to date has compared LLM-based augmentation against the embedding-space SMOTE-style retrieval (EmbSMOTE), a strong classical reference for imbalanced classification. In this study, a controlled benchmark of 11 augmentation methods, spanning classical perturbation, embedding-space retrieval, and LLM-based generation, is newly constructed on seven public text classification datasets covering class counts $K=2$-$28$ and imbalance ratios of 1.1 to over 500, evaluated with five random seeds per cell using macro F1, Welch's $t$-tests, five distributional metrics, and an LLM-family sensitivity analysis based on Qwen3-8B. The experimental results reveal that all LLM-based methods are statistically equivalent or inferior to EmbSMOTE, with the performance gap widening monotonically as imbalance increases and reaching $Δ\text{F1}_\text{macro}\!\approx\!0.063$ on GoEmotions-28. Furthermore, it is observed that surface-level uniqueness has negligible correlation with downstream performance, whereas LLM-specific artifacts, such as text elongation and label-distribution uniformization, are negatively associated with classification accuracy. Compared with six LLM-based and four classical augmentation baselines, these results demonstrate that the effective variable is not surface-level diversity but class-conditional structural fidelity, namely the degree to which augmented samples preserve the class-conditioned geometry of the training distribution. Accordingly, retrieval-based oversampling should be adopted as the default for imbalanced multi-class classification, and a higher empirical bar should be required before LLM-based augmentation is deployed in practice.

---

### Mimicry without understanding: the origins of decision bias in large language models

Authors: Eldad Yechiam, Adi Tarabeih
Published: 2026-06-03
Categories: cs.CL, cs.AI, cs.HC
arXiv: https://arxiv.org/abs/2608.12339v1

Large Language models (LLMs) were found to be susceptible to a host of social, affective, and cognitive biases. We examined two mechanisms through which such biases can be generated even when human preferences (in the training data) are not biased or when they are correctly categorized as being biased. The first is faulty mimicry of preferences based on human behavior: this involves LLMs inferring human preferences even when behaviors are logically unrelated to preferences. The second is mimicry of explicitly biased human behaviors. In four studies focusing on economic biases, we find that ChatGPT-4o and Qwen exhibited social proof biases even when prompted with reports of human behaviors that were clearly non-indicative of individuals' actual preferences. LLMs also displayed loss aversion when it was explicitly described as a bias. Indeed, when prompted with detailed scientific reports, the extent of the bias (i.e., loss aversion) in the scientific report predicted LLMs' own subsequent bias. Scientific papers of biases can thus become self-fulfilling prophecies, at least when it comes to LLMs' responses. The current study goes beyond fleshing out LLM biases and sheds light on the underlying component processes.

---

### SDAM: Structure-Difference-Aware Memory Evolution for Complex Text-to-SQL

Authors: Keyan Xu, Dingzirui Wang, Xuanliang Zhang, Qingfu Zhu, Wanxiang Che
Published: 2026-06-03
Categories: cs.CL
arXiv: https://arxiv.org/abs/2608.12338v1

Text-to-SQL aims to convert natural language questions into executable SQL queries. While memory-based agent system improves complex SQL generation, existing memory design neglect historical experience and suffer from weak structure analysis, shallow semantic understanding, and poor schema alignment. To address these challenges, we propose SDAM. Specifically, SDAM identifies potential errors via a structure-difference aware reasoning tree, extracts deep semantic rules through contradiction-aware reflection, and enhances structural consistency using a schema-grounded memory evolution mechanism to bind memory with database schemas. We integrate SDAM into a Text-to-SQL framework named SDAM-SQL. Experiment shows that SDAM-SQL achieves 2.0 and 0.4 improvement on BIRD-dev and Spider-test compared with mainstream Text-to-SQL methods, showing the effectiveness of SDAM-SQL.

---

### From Refuse to Richness: Rubric Rewards for Long-Form Hallucination Reinforcement Learning

Authors: Yudong Wang, Zhe Yang, Wenhan Ma, Rang Li, Qibin Yang, Weimin Xiong, Jiangshan Duo, Liang Zhao, Zhifang Sui
Published: 2026-06-03
Categories: cs.CL
arXiv: https://arxiv.org/abs/2608.12337v1

Rewards that penalize unsupported claims can improve grounding in long-form generation, but they can also teach models to answer less. We study this refusal-to-richness trade-off in long-form hallucination RL. Instead of using global richness proxies such as length, claim count, detail, or pairwise relevance, we represent each question with a key-point rubric that specifies the required and optional information a useful answer should cover. These rubrics define coverage directly and are used both for evaluation and as reward signals. Across grounding-only, proxy-based, rubric-only, and combined rewards, we find a stable trade-off: strict grounding rewards improve support but suppress coverage, while unconstrained rubric rewards improve coverage but weaken grounding. A soft combination of grounding, rubric coverage, and relevance gives the best balance in our experiments, improving in-distribution support while transferring better to out-of-distribution checklist tasks than either grounding-only or rubric-only rewards.

---

### StorySpark: Module-wise Evolutionary Search for Story Premise Generation

Authors: Yang Yang, Zining Zhong, Qian Cao, Jindong Li, Boyun Xu, Kaishen Yuan, Menglin Yang, Yutao Yue
Published: 2026-06-03
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2608.12336v1

A story premise is the creative spark from which a full narrative can grow. Yet LLM-based story generation has mostly emphasized later-stage planning, controllability, coherence, and prose expansion, while premise-level ideation remains comparatively underexplored. We introduce StorySpark, a module-wise evolutionary search framework for story premise generation. StorySpark operates over interpretable narrative modules such as background, persona, event, ending, and twist, treating each active module not as a static field to fill once, but as a local search space conditioned on the partial premise built so far. For each module, it generates alternatives, evaluates them in context, refines them through feedback-driven mutation and recombination, preserves complementary strengths with Pareto-guided selection, and reallocates frontier capacity to balance branch coverage with promising directions. Multi-view automatic and human evaluations show that StorySpark produces stronger final premises than competitive baselines, with especially consistent gains in originality; when expanded with the same story writer, its premises also lead to higher-quality downstream stories while maintaining completeness, fascination, and diverse usable narrative directions.

---

### HC-RAG: Evidence-Centric Retrieval-Augmented Generation over Heterogeneous Financial Filings

Authors: Siyuan Chen, Huaye Tan, You Li, Jiajun Liang
Published: 2026-06-03
Categories: cs.CL, cs.MM
arXiv: https://arxiv.org/abs/2608.12335v1

Financial question answering over annual reports requires more than retrieving semantically similar passages. It often involves identifying relevant companies and fiscal years, locating standardized filing sections, collecting textual and tabular evidence, and checking answers against the original documents. Existing RAG systems, however, usually flatten long filings into unordered chunks, pay limited attention to the typed structure of financial reports, and use fixed text-table fusion strategies without considering query intent. To address these limitations, we propose \textbf{HC-RAG}, a hierarchical cross-modal retrieval-augmented generation framework for evidence-centric financial QA. HC-RAG organizes filings into a typed financial evidence graph with documents, sections, text units, table units, and metadata nodes. It retrieves evidence through document-section-unit paths, aligns textual and tabular evidence in a shared retrieval space, and routes evidence according to four semantic intents: calculation, trend, fact, and comparison. We further introduce \textbf{Multi-Doc-2025}, a benchmark containing 2,327 expert-verified QA pairs from 179 SEC 10-K filings of 87 S\&P 500 companies across fiscal years 2022--2024, with labels for intent, difficulty, and structural evidence attributes. Experiments on public financial QA benchmarks and Multi-Doc-2025 show that HC-RAG improves both answer quality and evidence localization, especially in long-document, table-related, and cross-document settings. HC-RAG outperforms RAPTOR by 6.6 F1 points on DocFinQA and GraphRAG by 10.9 F1 points on Multi-Doc-2025. Evidence-level analysis and ablation studies show that the improvements mainly come from more accurate section localization, table grounding, cross-document evidence aggregation, and intent-aware text-table routing.

---

### Marginally Useful: An Information-Gap Identity in Conformal Prediction

Authors: Peter Cotton
Published: 2026-06-03
Categories: q-fin.ST, math.ST, q-fin.MF, stat.ML
arXiv: https://arxiv.org/abs/2608.07479v2

Conformal prediction has been touted as a more formal, rigorous approach to adding uncertainty to a forecast. The sole objective of this note is to point out that rigor cuts both ways in the case of residual pooling, the technique used in the vast majority of conformal prediction applications. The fact that unconditional guarantee of coverage is provided is not in question, but we make clear, we believe for the first time, that there is an opposing guarantee too: a permanent gambit of logarithmic-score regret which no amount of data or tuning can subsequently reduce. We give the exact size of the sacrifice, and a financial reading of it as the growth rate of an oracle adversary betting against odds set by someone using residual pooling.

---

### Schema-Aware Localisation (SAL): Live Schema Grounding and Hallucination Validation for Oracle NL2SQL

Authors: Sanjay Mishra, Divya Chukkapalli, Ganesh R. Naik
Published: 2026-06-03
Categories: cs.AI
arXiv: https://arxiv.org/abs/2607.22572v1

Large language models can generate fluent SQL from natural language, but on real enterprise Oracle databases they frequently fail at execution time: columns and aliases are hallucinated and dialect-specific syntax is missed, leading to ORA-00904 invalid-identifier errors. In this setting, failures are primarily due to missing schema grounding: the model cannot know which tables and columns actually exist. This paper introduces Schema-Aware Localisation (SAL), a lightweight middleware layer for Oracle NL2SQL that requires no model retraining. SAL queries Oracle's USER_TAB_COLUMNS catalog to build a live schema map, selects a relevant table subset for each question (falling back to the full schema for multi-table queries), and injects this ground-truth context into the LLM prompt. Generated SQL is then checked by the Hallucination Index (Hidx), which validates every alias.column reference against the live catalog, automatically rewrites predictable prefix errors, and otherwise triggers a structured retry with itemised corrections. We evaluate SAL on 500 TPC-H natural language questions executed against a live Oracle Autonomous Database 23c instance using GPT-4o-mini. Without any schema grounding, execution-grounded truth (EGT; executes and matches the reference result set) is 2.2% (12/500). A hand-written static schema hint brings EGT to 62.0%. SAL, with no manual schema curation, achieves 62.6% EGT (96% simple, 95% medium, 40.7% complex) while reducing execution failures from 97.6% to 2.6%.

---

### Stateful Guardrails for Multi-Turn LLM Systems: A Conversational Risk Accumulation Framework

Authors: Sanjay Mishra, Divya Chukkapalli, Ganesh R. Naik
Published: 2026-06-03
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2607.19361v1

Most safety guardrails for large language models (LLMs) evaluate each prompt-response pair in isolation, which misses failures that arise only over a dialogue as benign turns compose into harm. We term this Conversational Risk Accumulation (CRA): gradual intent drift, fragmented assembly of prohibited instructions, and sensitivity build-up from repeated disclosures. We propose a session-layer CRA Framework that tracks three trajectory signals: semantic drift from a session anchor, a sensitivity-weighted information accumulation graph over extracted entities, and a compliance-gradient signal capturing increasing willingness to comply. For scoring, we provide (i) an unsupervised convex fusion for attribution and ablations, and (ii) CRA-Net DA, a compact learned trajectory model trained with family-adversarial objectives to reduce length and topic-coverage confounds. To benchmark CRA, we release CRA-Bench v0.1 (1,200 eight-turn sessions across three threat families with topic-matched benign twins), CRA-Bench v0.2 (LLM-paraphrased variants to reduce template artifacts), and an extended 5-family set (2,000 sessions adding persona priming and context stuffing). We introduce a trajectory-native evaluation protocol with session-level splits, mixed-set threshold calibration, Trajectory AUROC, turns-to-detection, calibrated false-positive metrics, bootstrap confidence intervals, leave-one-family-out diagnostic stress tests, and synthetic-to-human transfer checks. Claims focus on within-distribution session scoring on CRA-Bench and human-transfer subsets.

---

### How Indian Dermatologists are Utilizing Artificial Intelligence for Clinical Practice and Workflow Management: A Nationwide Survey with a Special Focus on atopic dermatitis

Authors: Dipayan Sengupta, Saumya Panda, Sandipan Dhar, Dipankar De, Deepika Pandhi, Narayanan B
Published: 2026-06-03
Categories: cs.CY, cs.AI
arXiv: https://arxiv.org/abs/2607.01252v1

Background: Dermatology AI has mainly focused on image-based diagnosis, while chronic disease workflows have received less attention. We surveyed Indian dermatologists to map routine clinical challenges, with a focus on atopic dermatitis (AD), and assess current AI use. Methods: A nationwide cross-sectional survey commissioned by the Society for Eczema Studies included 377 practicing Indian dermatologists. The survey assessed clinical challenges, AD workflow barriers, AI use, adoption barriers, and ethical concerns. Analyses used descriptive statistics, chi-square tests, false discovery rate correction, and multivariable logistic regression. Results: Patient adherence (61.3%) and treatment planning in difficult or refractory cases (57.0%) were reported more often than diagnostic uncertainty (48.0%). In AD care, severity scoring was reported as a challenge by 47.7% and had the lowest satisfaction among measured workflow areas. Current AI use was reported by 49.9%, most often involving general large language models for literature synthesis, documentation, and academic tasks rather than specialized image analysis. Barriers differed by experience: dermatologists with more than 20 years of practice more often cited lack of training, while those with 5 years or less more often cited lack of clinical utility after trying AI tools. AI users were more likely than non-users to report concern about patient self-misdiagnosis and anxiety, which remained significant after adjustment for experience and academic affiliation. Conclusion: Respondents reported using general-purpose AI mainly for cognitive and administrative tasks, while their clinical needs centered on chronic disease management and AD workflow support. Clinician-supervised workflow tools may be more useful than standalone diagnostic applications.

---

### UniCA: Bi-directional Cross-Attention with Positive Similarity Loss for Robust Multi-Modal Retrieval

Authors: Yini Huang, Wenlong Zhang
Published: 2026-06-03
Categories: cs.IR, cs.CV
arXiv: https://arxiv.org/abs/2606.28350v1

Multi-modal retrieval has become increasingly critical for handling the growing volume of integrated visual-textual data in real-world applications, but existing frameworks rely on implicit fusion via text encoder self-attention, limiting explicit cross-modal semantic alignment. To address this gap, this paper proposes UniCA (Unified Cross-Attention Encoder), a multi-modal retrieval model with four key innovations: 1) a bi-directional cross-attention (Bi-CA) block that enables active semantic exchange between visual and textual tokens prior to concatenation, capturing inter-modal correlations more efficiently. 2) a Positive Similarity Loss that optimizes absolute semantic proximity between query and positive candidate embeddings. 3) a streamlined dataset UMR-S10 (Universal Multimodal Retrieval Sample 10%) to reduce computational costs while retaining semantic diversity and task representativeness. 4) an experimental validation on the WebQA benchmark demonstrates that UniCA outperforms the baseline model across Hybrid and Image-Text tasks, achieving improvements of up to 4.09% in Recall@5, 3.28% in Recall@10, and 3.96% in MRR@1 for the hybrid task. UniCA provides an efficient and robust solution for multi-modal retrieval, lowering deployment barriers through its lightweight dataset and enhanced fusion mechanism.

---

### HMARS: A Hierarchical Multi-Agent Memory System for Long-Context Reasoning

Authors: Zeju Li, Ziyang Zheng, Yizhou Zhou, Qiang Xu
Published: 2026-06-03
Categories: cs.IR, cs.AI
arXiv: https://arxiv.org/abs/2606.28349v1

Long-context reasoning requires models to access, retrieve, and integrate evidence scattered across documents, dialogues, and accumulated interaction histories. Standard retrieval-augmented generation reduces this problem to top-$K$ chunk retrieval, but such passive access can discard relevant evidence before reasoning begins, especially when relevance depends on broader context. We propose HMARS, a hierarchical multi-agent memory system that treats long contexts as managed memory rather than a flat retrieval corpus. Sub-agents maintain grounded access to bounded memory regions, mid-agents manage regional context and provide query-specific coordination, and a frontier model performs final reasoning over retrieved evidence pages. To evaluate this view, we construct two diagnostic benchmarks targeting evidence breadth and context-dependent relevance. Across long-document and multi-turn memory tasks, HMARS achieves the best overall performance against retrieval, reranking, full-context, graph-based, and agentic long-context baselines. Evidence coverage analysis further shows that its gains come from retrieving the required supporting evidence more completely, rather than merely changing the final answer prompt.

---

### A Geometry-Informed Computer Vision Method for Detecting and Examining Overtaking Vehicles From A Bicycle

Authors: Gandhimathi Padmanaban, Rayane Moustafa, Fred Feng
Published: 2026-06-03
Categories: cs.CV, cs.HC
arXiv: https://arxiv.org/abs/2606.23699v1

Instrumented bicycle studies have produced direct field evidence on vehicle passing behavior, but extracting overtaking events from continuous rear-facing video has remained dependent on manual, frame-by-frame annotation. This bottleneck constrains sample sizes and limits naturalistic cycling safety research. We present a geometry-informed computer vision pipeline that automates overtaking event detection from a single bicycle-mounted camera without multi-sensor configurations or explicit camera calibration. The system combines RT-DETR object detection with ByteTrack multi-object tracking through a three-stage geometric validation module enforcing bearing angle trend, apparent size growth, and spatial confirmation criteria derived from perspective projection principles. Validated on 315 manually annotated real-world overtaking events from urban roads in Ann Arbor, Michigan, the pipeline achieved 97.8% recall with zero false positives. The system identified overtaking intentions a mean of 2.44 seconds before vehicle passage, with 84.1% of events exceeding the 1.5-second human reaction time threshold, demonstrating feasibility for active cyclist warning. Lateral passing distance measurements from 96 events revealed 33.3% of passes below the 5-foot (152.4 cm) threshold, consistent with non-compliance rates in prior field and self-reported studies. A preliminary calibration-free lateral distance estimation approach using bounding box geometric features achieved mean absolute errors of 13-14 cm under leave-one-out cross-validation, sufficient to distinguish close passes from standard passes for safety categorization. By automating event isolation from consumer-grade footage, the system removes the primary annotation bottleneck of instrumented bicycle research and provides a scalable foundation for vehicle-bicycle interaction analysis across larger datasets and diverse urban environments.

---

### Hypothesis-Disciplined Multi-Agent Automated Formalization of Asymptotic Statistical Theory

Authors: Tingzhou Wei, Zeyu Zheng, Ethan X. Fang, Junwei Lu
Published: 2026-06-03
Categories: cs.AI, cs.LG, cs.LO, math.ST
arXiv: https://arxiv.org/abs/2606.20642v1

Asymptotic statistical theory is a challenging domain for AI-assisted formalization: its central results mix convergence statements, asymptotic expansions, functional analysis, and regularity conditions that have a large gap from existing infrastructure in Lean 4 formalization. To address these challenges, we propose a hypothesis-disciplined Lean 4 formalization pipeline built from multiple agents: a manager that coordinates seven specialist roles for proof planning, skeleton scaffolding, Mathlib reconnaissance, proof construction, integration, independent review, and audit. The main methodological discipline is the hypothesis-disciplined audit, implemented by the Auditor agent: every main-theorem hypothesis and concept-layer field must be anchored in the source mathematical prose, justified as a Lean encoding adapter, marked as source-implied, or rejected as an unsupported strengthening. Using this workflow, we build a systematic formalization of asymptotic statistical theory, especially the parametric and semi-parametric models' asymptotic distribution and efficiency results. The resulting Lean development is axiom-clean and source-faithful, with Lean-checked and human-audited proofs of core parametric and semi-parametric theorems organized so that theorem-agnostic infrastructure and statistical concept definitions are separated from theorem-specific assembly. The formalization results are available at https://github.com/junwei-lu/Lean-Asymptotic-Statistical-Theory.

---

### Towards Multi-Agent-Simulation-Based Community Note Evaluation

Authors: Changxi Wen, Shuning Zhang, Bohao Chu, Yuwei Chuai, Hui Wang, Dai Shi, Xin Yi, Hewu Li
Published: 2026-06-03
Categories: cs.SI, cs.AI
arXiv: https://arxiv.org/abs/2606.18268v1

Community-based fact-checking that relies on cross-consensus is expanding rapidly on social media platforms. However, the delay and low-ratio of cross-consensus community fact-checks rated by human contributors remains a significant challenge. To address this, we first created ComRate, a large-scale dataset comprising 2.5 million community notes and over 209 million ratings sourced from $\mathbb{X}$. We then propose MultiCom, a persona-guided multi-agent rating framework for community note evaluation. MultiCom simulates diverse rater population by clustering contributors in a matrix-factorized rater space and prompting persona agents to generate structured assessments based on the official community notes rating schema. These agents output structured and explainable judgments, such as confidence, agreement signals and reasons. An out-of-fold calibrated aggregation algorithm combines features such as raw votes and diagnostic reason signals for reliable prediction. Extensive evaluations demonstrate that MultiCom outperforms alternative methods, achieving an average accuracy of 84.7% (balanced accuracy 68.3%, macro-F1 60.1%) on the evaluation set.

---

### RadSEM: A Finding-by-Finding Metric for Clinical Consistency in Radiology Reports

Authors: Zhenhong Yang, Zhuoyun Liu, Jintao Fei, Wen Tang, Shichao Quan, Jun Zhao, Jun Xu
Published: 2026-06-03
Categories: q-bio.QM, cs.LG
arXiv: https://arxiv.org/abs/2606.17062v1

Radiology report evaluation must distinguish clinical compatibility from surface similarity, because negation, laterality, or normal-abnormal polarity can reverse a finding. We propose RadSEM (Radiology Sentence-Level Evaluation Metric), a constrained LLM-assisted metric for reference-based evaluation of radiology Findings. RadSEM rewrites reference and generated reports into ordered atomic finding sentences, each expressing one site-finding proposition. It then performs contradiction-constrained many-to-many matching: incompatible pairs such as "effusion" and "no effusion" receive no credit, while compatible granularity differences can receive partial credit. A deterministic stage weights pairs by part-whole and abnormal-detail relationships, counts unmatched findings, and produces an abnormal-focused weighted F1 score. Thus, the LLM supports structured rewriting and local alignment rather than acting as an opaque judge. We evaluate RadSEM with SSREE, a controlled monotonicity stress test built from 2,448 de-identified reports expanded into five graded corruption levels. RadSEM achieves Kendall tau_b of 0.957, all-pairs concordance of 97.8%, adjacent concordance of 95.0%, and strict five-level ordering for 81.9% of reports, outperforming radiology-specific and general text metrics while avoiding the failure in which polarity-inverted reports regain lexical overlap. On the same SSREE set, RadSEM outperforms the Ref-anchored RadSEM-Alt policy, improving adjacent concordance from 90.7% to 95.0% and strict ordering from 67.2% to 81.9%. On a 599-triplet synonym/antonym subset, RadSEM prefers synonyms in 597 cases (99.67%). These results suggest that explicit finding units, contradiction-aware matching, and abnormal-focused deterministic scoring make report scoring more interpretable and sensitive to clinically meaningful errors. Code is available at https://github.com/jdh-algo/RadSEM.

---

### Do Large Language Models Have Emotions?

Authors: Amit Goldenberg, James J. Gross
Published: 2026-06-03
Categories: q-bio.NC, cs.AI, cs.HC
arXiv: https://arxiv.org/abs/2606.14742v1

Do LLMs have emotions? A recent paper from Anthropic reports finding internal representations of emotion concepts in Claude Sonnet 4.5, concluding that the LLM has 'functional emotions.' We evaluate this claim against what is known about how emotions actually function in biological systems. We argue that emotions serve two core functions: the context-sensitive interpretation of situations, and the reorganization of processing across multiple systems in response to those interpretations. The Anthropic findings offer partial support for the first function, though the consistent, discrete emotional representations identified in Claude sit uneasily with affective neuroscience findings that human emotion is characterized by variable rather than uniform neural signatures. On the second function, the evidence is mixed: Claude's representations modulate output without producing the dynamic reorganization of attention, decision speed, and motivational state that defines emotion in biological systems. We close by proposing what it would take for an LLM to have emotions.

---

### ProHiFlo: Hierarchical Flow Matching with Functional Guidance for De Novo Protein Generation

Authors: Chuanzhen Wang, Meade Cleti, Pete Jano
Published: 2026-06-03
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2606.11243v1

De novo protein generation has transformative potential in therapeutic design, enzyme engineering, and synthetic biology. While diffusion-based and flow matching approaches have achieved progress, they typically operate at single resolution and lack mechanisms for incorporating functional constraints. We introduce ProHiFlo, a hierarchical flow matching framework with three innovations: (1) coarse-to-fine generation that models backbone geometry before refining to all-atom coordinates, reducing computational cost while maintaining accuracy; (2) functional guidance leveraging pretrained predictors to steer generation toward desired properties without retraining; (3) adaptive SE(3)-equivariant architecture for efficient multi-scale processing. Experiments on unconditional generation, motif scaffolding, and functional design demonstrate state-ofthe-art performance while requiring 4 fewer sampling steps. On enzyme active site scaffolding, ProHiFlo achieves 58.9% success rate compared to 41.2% for RFDiffusion.

---

### Tractogram foundation model

Authors: Guikun Chen, Yuqian Chen, Yijie Li, Yogesh Rathi, Nikos Makris, Fan Zhang, Wenguan Wang, Lauren J. O'Donnell
Published: 2026-06-03
Categories: eess.IV, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2606.09893v1

Diffusion MRI (dMRI) tractography is the only noninvasive approach for mapping white-matter pathways in the living human brain. It represents each brain as a tractogram: a large, unordered set of three-dimensional streamlines that includes information about both local streamline geometry and whole-brain anatomical organization. This structure makes tractograms a natural but challenging target for representation learning. Existing methods treat streamline classification and subject-level prediction as separate problems: streamline classifiers focus on geometric patterns, whereas subject-level prediction often depends on hand-crafted features. As a result, current methods do not learn reusable representations that connect streamline anatomy with whole-brain inter-subject variation. Here we introduce TractFM, a tractogram foundation model that learns reusable representations directly from whole-brain streamline sets. TractFM combines a local streamline encoder with a permutation-equivariant tractogram encoder, allowing all streamlines from a subject to be contextualized jointly in a single forward pass. Pretraining on dense anatomical tract parcellation, i.e., assigning anatomical labels to individual streamlines, yields two complementary representations: contextualized streamline-level embeddings for tract parcellation and compact subject-level descriptors for downstream prediction of subject phenotypes. Across three tractography algorithms and five dMRI datasets, TractFM transfers to both streamline-level and subject-level tasks. Its frozen representations achieve accurate tract parcellation and predict age and sex across independent datasets. These results show that whole-brain geometric context, learned once, can generalize across tractography pipelines, datasets, and prediction tasks.

---