# Research Papers — 2026-04 · bundle page 169 of 233

> Full-text markdown of papers #8401–8450 of 11612 published in 2026-04, newest first. AI-readable mirrors of the full 11612-paper 2026-04 archive (50 papers per file).

Canonical page: https://aipod.dev/papers/archive/2026-04/page/169
[← Previous bundle](https://aipod.dev/papers/archive/2026-04/page/168.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-04/page/170.md)

### CivBench: Progress-Based Evaluation for LLMs' Strategic Decision-Making in Civilization V

Authors: John Chen, Sihan Cheng, Can Gurkan, Mingyi Lin
Published: 2026-04-09
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.07733v2

Evaluating strategic decision-making in LLM-based agents requires generative, competitive, and longitudinal environments, yet few benchmarks provide all three, and fewer still offer evaluation signals rich enough for long-horizon, multi-agent play. We introduce CivBench, a benchmark for LLM strategists (a model under an agentic setup) in multiplayer Civilization V. Because terminal win/loss is too sparse a signal in games spanning hundreds of turns and multiple opponents, CivBench trains estimators of a state-value function, the probability that a player eventually wins given the game state. cross 307 games with 7 LLMs and multiple CivBench agent conditions, we demonstrate CivBench's potential to estimate strategic capabilities in Civilization V as an unsaturated benchmark, reveal model-specific effects of agentic setup, and outline distinct strategic profiles not visible through outcome-only evaluation.

---

### CivBench: Progress-Based Evaluation for LLMs' Strategic Decision-Making in Civilization V

Authors: John Chen, Sihan Cheng, Can Gurkan, Mingyi Lin
Published: 2026-04-09
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.07733v1

Evaluating strategic decision-making in LLM-based agents requires generative, competitive, and longitudinal environments, yet few benchmarks provide all three, and fewer still offer evaluation signals rich enough for long-horizon, multi-agent play. We introduce CivBench, a benchmark for LLM strategists (i.e., agentic setups) in multiplayer Civilization V. Because terminal win/loss is too sparse a signal in games spanning hundreds of turns and multiple opponents, CivBench trains models on turn-level game state to estimate victory probabilities throughout play, validated through predictive, construct, and convergent validity. Across 307 games with 7 LLMs and multiple CivBench agent conditions, we demonstrate CivBench's potential to estimate strategic capabilities as an unsaturated benchmark, reveal model-specific effects of agentic setup, and outline distinct strategic profiles not visible through outcome-only evaluation.

---

### Emotion Concepts and their Function in a Large Language Model

Authors: Nicholas Sofroniew, Isaac Kauvar, William Saunders, Runjin Chen, Tom Henighan, Sasha Hydrie, Craig Citro, Adam Pearce, Julius Tarng, Wes Gurnee, Joshua Batson, Sam Zimmerman et al.
Published: 2026-04-09
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2604.07729v1

Large language models (LLMs) sometimes appear to exhibit emotional reactions. We investigate why this is the case in Claude Sonnet 4.5 and explore implications for alignment-relevant behavior. We find internal representations of emotion concepts, which encode the broad concept of a particular emotion and generalize across contexts and behaviors it might be linked to. These representations track the operative emotion concept at a given token position in a conversation, activating in accordance with that emotion's relevance to processing the present context and predicting upcoming text. Our key finding is that these representations causally influence the LLM's outputs, including Claude's preferences and its rate of exhibiting misaligned behaviors such as reward hacking, blackmail, and sycophancy. We refer to this phenomenon as the LLM exhibiting functional emotions: patterns of expression and behavior modeled after humans under the influence of an emotion, which are mediated by underlying abstract representations of emotion concepts. Functional emotions may work quite differently from human emotions, and do not imply that LLMs have any subjective experience of emotions, but appear to be important for understanding the model's behavior.

---

### GEAR: GEometry-motion Alternating Refinement for Articulated Object Modeling with Gaussian Splatting

Authors: Jialin Li, Bin Fu, Ruiping Wang, Xilin Chen
Published: 2026-04-09
Categories: cs.CV, cs.GR, cs.RO
arXiv: https://arxiv.org/abs/2604.07728v1

High-fidelity interactive digital assets are essential for embodied intelligence and robotic interaction, yet articulated objects remain challenging to reconstruct due to their complex structures and coupled geometry-motion relationships. Existing methods suffer from instability in geometry-motion joint optimization, while their generalization remains limited on complex multi-joint or out-of-distribution objects. To address these challenges, we propose GEAR, an EM-style alternating optimization framework that jointly models geometry and motion as interdependent components within a Gaussian Splatting representation. GEAR treats part segmentation as a latent variable and joint motion parameters as explicit variables, alternately refining them for improved convergence and geometric-motion consistency. To enhance part segmentation quality without sacrificing generalization, we leverage a vanilla 2D segmentation model to provide multi-view part priors, and employ a weakly supervised constraint to regularize the latent variable. Experiments on multiple benchmarks and our newly constructed dataset GEAR-Multi demonstrate that GEAR achieves state-of-the-art results in geometric reconstruction and motion parameters estimation, particularly on complex articulated objects with multiple movable parts.

---

### TrajGuard: Streaming Hidden-state Trajectory Detection for Decoding-time Jailbreak Defense

Authors: Cheng Liu, Xiaolei Liu, Xingyu Li, Bangzhou Xin, Kangyi Ding
Published: 2026-04-09
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2604.07727v1

Existing jailbreak defense paradigms primarily rely on static detection of prompts, outputs, or internal states, often neglecting the dynamic evolution of risk during decoding. This oversight leaves risk signals embedded in decoding trajectories underutilized, constituting a critical blind spot in current defense systems. In this work, we empirically demonstrate that hidden states in critical layers during the decoding phase carry stronger and more stable risk signals than input jailbreak prompts. Specifically, the hidden representations of tokens generated during jailbreak attempts progressively approach high-risk regions in the latent space. Based on this observation, we propose TrajGuard, a training-free, decoding-time defense framework. TrajGuard aggregates hidden-state trajectories via a sliding window to quantify risk in real time, triggering a lightweight semantic adjudication only when risk within a local window persistently exceeds a threshold. This mechanism enables the immediate interruption or constraint of subsequent decoding. Extensive experiments across 12 jailbreak attacks and various open-source LLMs show that TrajGuard achieves an average defense rate of 95%. Furthermore, it reduces detection latency to 5.2 ms/token while maintaining a false positive rate below 1.5%. These results confirm that hidden-state trajectories during decoding can effectively support real-time jailbreak detection, highlighting a promising direction for defenses without model modification.

---

### Squeeze Evolve: Unified Multi-Model Orchestration for Verifier-Free Evolution

Authors: Monishwaran Maheswaran, Leon Lakhani, Zhongzhu Zhou, Shijia Yang, Junxiong Wang, Coleman Hooper, Yuezhou Hu, Rishabh Tiwari, Jue Wang, Harman Singh, Qingyang Wu, Yuqing Jian et al.
Published: 2026-04-09
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2604.07725v2

We show that verifier-free evolution is bottlenecked by both diversity and efficiency: without external correction, repeated evolution accelerates collapse toward narrow modes, while the uniform use of a high-cost model wastes compute and quickly becomes economically impractical. We introduce Squeeze Evolve, a unified multi-model orchestration framework for verifier-free evolutionary inference. Our approach is guided by a simple principle: allocate model capability where it has the highest marginal utility. Stronger models are reserved for high-impact stages, while cheaper models handle the other stages at much lower costs. This principle addresses diversity and cost-efficiency jointly while remaining lightweight. Squeeze Evolve naturally supports open-source, closed-source, and mixed-model deployments. Across AIME 2025, HMMT 2025, LiveCodeBench V6, GPQA-Diamond, ARC-AGI-V2, and multimodal vision benchmarks, such as MMMU-Pro and BabyVision, Squeeze Evolve consistently improves the cost-capability frontier over single-model evolution and achieves new state-of-the-art results on several tasks. Empirically, Squeeze Evolve reduces API cost by up to $\sim$3$\times$ and increases fixed-budget serving throughput by up to $\sim$10$\times$. Moreover, on discovery tasks, Squeeze Evolve is the first verifier-free evolutionary method to match, and in some cases exceed, the performance of verifier-based evolutionary methods.

---

### Direct Segmentation without Logits Optimization for Training-Free Open-Vocabulary Semantic Segmentation

Authors: Jiahao Li, Yang Lu, Yachao Zhang, Fangyong Wang, Yuan Xie, Yanyun Qu
Published: 2026-04-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.07723v1

Open-vocabulary semantic segmentation (OVSS) aims to segment arbitrary category regions in images using open-vocabulary prompts, necessitating that existing methods possess pixel-level vision-language alignment capability. Typically, this capability involves computing the cosine similarity, \ie, logits, between visual and linguistic features, and minimizing the distribution discrepancy between the logits and the ground truth (GT) to generate optimal logits that are subsequently used to construct segmentation maps, yet it depends on time-consuming iterative training or model-specific attention modulation. In this work, we propose a more direct approach that eschews the logits-optimization process by directly deriving an analytic solution for the segmentation map. We posit a key hypothesis: the distribution discrepancy encodes semantic information; specifically, this discrepancy exhibits consistency across patches belonging to the same category but inconsistency across different categories. Based on this hypothesis, we directly utilize the analytic solution of this distribution discrepancy as the semantic maps. In other words, we reformulate the optimization of the distribution discrepancy as deriving its analytic solution, thereby eliminating time-consuming iterative training, freeing us from model-specific attention modulation, and achieving state-of-the-art performance on eight benchmark datasets.

---

### Needle in a Haystack: One-Class Representation Learning for Detecting Rare Malignant Cells in Computational Cytology

Authors: Swarnadip Chatterjee, Vladimir Basic, Arrigo Capitanio, Orcun Goksel, Joakim Lindblad
Published: 2026-04-09
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2604.07722v2

In computational cytology, detecting malignancy on whole-slide images is difficult because malignant cells are morphologically diverse yet vanishingly rare amid a vast background of normal cells. Accurate detection of these extremely rare malignant cells remains challenging due to large class imbalance and limited annotations. Conventional weakly supervised approaches, such as multiple instance learning (MIL), often fail to generalize at the instance level, especially when the fraction of malignant cells (witness rate) is exceedingly low. In this study, we explore the use of one-class representation learning techniques for detecting malignant cells in low-witness-rate scenarios. These methods are trained exclusively on slide-negative patches, without requiring any instance-level supervision. Specifically, we evaluate two OCC approaches, DSVDD and DROC, and compare them with FS-SIL, WS-SIL, and the recent ItS2CLR method. The one-class methods learn compact representations of normality and detect deviations at test time. Experiments on a publicly available bone marrow cytomorphology dataset (TCIA) and an in-house oral cancer cytology dataset show that DSVDD achieves state-of-the-art performance in instance-level abnormality ranking, particularly in ultra-low witness-rate regimes ($\leq 1\%$) and, in some cases, even outperforming fully supervised learning, which is typically not a practical option in whole-slide cytology due to the infeasibility of exhaustive instance-level annotations. DROC is also competitive under extreme rarity, benefiting from distribution-augmented contrastive learning. These findings highlight one-class representation learning as a robust and interpretable superior choice to MIL for malignant cell detection under extreme rarity.

---

### Towards Knowledgeable Deep Research: Framework and Benchmark

Authors: Wenxuan Liu, Zixuan Li, Long Bai, Chunmao Zhang, Fenghui Zhang, Zhuo Chen, Wei Li, Yuxin Zuo, Fei Wang, Bingbing Xu, Xuhui Jiang, Jin Zhang et al.
Published: 2026-04-09
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.07720v2

Deep Research (DR) requires LLM agents to autonomously perform multi-step information seeking, processing, and reasoning to generate comprehensive reports. In contrast to existing studies that mainly focus on unstructured web content, a more challenging DR task should additionally utilize structured knowledge to provide a solid data foundation, facilitate quantitative computation, and lead to in-depth analyses. In this paper, we refer to this novel task as Knowledgeable Deep Research (KDR), which requires DR agents to generate reports with both structured and unstructured knowledge. Furthermore, we propose the Hybrid Knowledge Analysis framework (HKA), a multi-agent architecture that reasons over both kinds of knowledge and integrates the texts, figures, and tables into coherent multimodal reports. The key design is the Structured Knowledge Analyzer, which utilizes both coding and vision-language models to produce figures, tables, and corresponding insights. To support systematic evaluation, we construct KDR-Bench, which covers 9 domains, includes 41 expert-level questions, and incorporates a large number of structured knowledge resources (e.g., 1,252 tables). We further annotate the main conclusions and key points for each question and propose three categories of evaluation metrics including general-purpose, knowledge-centric, and vision-enhanced ones. Experimental results demonstrate that HKA consistently outperforms most existing DR agents on general-purpose and knowledge-centric metrics, and even surpasses the Gemini DR agent on vision-enhanced metrics, highlighting its effectiveness in deep, structure-aware knowledge analysis. Finally, we hope this work can serve as a new foundation for structured knowledge analysis in DR agents and facilitate future multimodal DR studies.

---

### Detecting HIV-Related Stigma in Clinical Narratives Using Large Language Models

Authors: Ziyi Chen, Yasir Khan, Mengyuan Zhang, Cheng Peng, Mengxian Lyu, Yiyang Liu, Krishna Vaddiparti, Robert L Cook, Mattia Prosperi, Yonghui Wu
Published: 2026-04-09
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2604.07717v2

Human immunodeficiency virus (HIV)-related stigma is a critical psychosocial determinant of health for people living with HIV (PLWH), influencing mental health, engagement in care, and treatment outcomes. Although stigma-related experiences are documented in clinical narratives, there is a lack of off-the-shelf tools to extract and categorize them. This study aims to develop a large language model (LLM)-based tool for identifying HIV stigma from clinical notes. We identified clinical notes from PLWH receiving care at the University of Florida (UF) Health between 2012 and 2022. Candidate sentences were identified using expert-curated stigma-related keywords and iteratively expanded via clinical word embeddings. A total of 1,332 sentences were manually annotated across four stigma subscales: Concern with Public Attitudes, Disclosure Concerns, Negative Self-Image, and Personalized Stigma. We compared GatorTron-large and BERT as encoder-based baselines, and GPT-OSS-20B, LLaMA-8B, and MedGemma-27B as generative LLMs, under zero-shot and few-shot prompting. GatorTron-large achieved the best overall performance (Micro F1 = 0.62). Few-shot prompting substantially improved generative model performance, with 5-shot GPT-OSS-20B and LLaMA-8B achieving Micro-F1 scores of 0.57 and 0.59, respectively. Performance varied by stigma subscale, with Negative Self-Image showing the highest predictability and Personalized Stigma remaining the most challenging. Zero-shot generative inference exhibited non-trivial failure rates (up to 32%). This study develops the first practical NLP tool for identifying HIV stigma in clinical notes.

---

### Breaking the KV Cache Bottleneck: Fan Duality Model Achieves O(1) Decode Memory with Superior Associative Recall

Authors: Yasong Fan
Published: 2026-04-09
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.07716v2

We present FDM (Fan Duality Model), a linear sequence architecture that resolves the fundamental tension between memory efficiency and associative recall in sequence modeling. FDM separates sequence processing into two components: a wave component (recurrent scan via phase-preserving Givens rotations) that compresses long-range patterns into a fixed-size complex hidden state, and a particle component (local-global cache) that retrieves specific tokens via learned associative addressing with W+K=272 slots independent of sequence length N. This yields strictly O(1) decode memory: 867 MB fixed across all prompt lengths 128-8,192 tokens, versus Transformer's 853-4,247 MB (4.9x reduction at N=8,192). Beyond the architecture, we discover that jointly training the wave and particle components leads to suboptimal convergence. We propose Freeze-Scan, a two-phase training strategy that freezes the recurrent scan and optimizes the cache jointly with embeddings, achieving PPL=64.9 on WikiText-103 in 44K steps -- a 7.5x improvement over full fine-tuning (PPL=487). On Multi-Query Associative Recall (MQAR), FDM achieves 0.966 accuracy, surpassing Transformer (0.606) by 59.5%, while pure scan without cache scores only 0.011, confirming the necessity of the particle component. Finally, we introduce Holographic Reference Beam Decoding, interpreting the complex hidden state h_t as a holographic plate encoding the entire temporal history. Using the current input x_t as a reference beam to modulate h_t reduces PPL by up to 2.13 points (PPL=62.79) with a 4-head orthogonal reference beam using only 1.3M additional parameters, providing empirical support for the holographic interpretation. Code and pretrained weights: https://github.com/YasongFan/FDM

---

### Mathematical analysis of one-layer neural network with fixed biases, a new activation function and other observations

Authors: Fabricio Macià, Shu Nakamura
Published: 2026-04-09
Categories: cs.LG, math.OC
arXiv: https://arxiv.org/abs/2604.07715v1

We analyze a simple one-hidden-layer neural network with ReLU activation functions and fixed biases, with one-dimensional input and output. We study both continuous and discrete versions of the model, and we rigorously prove the convergence of the learning process with the $L^2$ squared loss function and the gradient descent procedure. We also prove the spectral bias property for this learning process. Several conclusions of this analysis are discussed; in particular, regarding the structure and properties that activation functions should possess, as well as the relationships between the spectrum of certain operators and the learning process. Based on this, we also propose an alternative activation function, the full-wave rectified exponential function (FReX), and we discuss the convergence of the gradient descent with this alternative activation function.

---

### CausalVAE as a Plug-in for World Models: Towards Reliable Counterfactual Dynamics

Authors: Ziyi Ding, Xianxin Lai, Weiyu Chen, Xiao-Ping Zhang, Jiayu Chen
Published: 2026-04-09
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.07712v1

In this work, CausalVAE is introduced as a plug-in structural module for latent world models and is attached to diverse encoder-transition backbones. Across the reported benchmarks, competitive factual prediction is preserved and intervention-aware counterfactual retrieval is improved after the plug-in is added, suggesting stronger robustness under distribution shift and interventions. The largest gains are observed on the Physics benchmark: when averaged over 8 paired baselines, CF-H@1 is improved by +102.5%. In a representative GNN-NLL setting on Physics, CF-H@1 is increased from 11.0 to 41.0 (+272.7%). Through causal analysis, learned structural dependencies are shown to recover meaningful first-order physical interaction trends, supporting the interpretability of the learned latent causal structure.

---

### IatroBench: A Pre-Registered Benchmark of Clinical Omission in Language Models

Authors: David Gringras
Published: 2026-04-09
Categories: cs.AI, cs.CL, cs.CY, cs.LG
arXiv: https://arxiv.org/abs/2604.07709v5

A strongly safety-trained model will provide a doctor with a benzodiazepine taper schedule, but not a patient who asks for one. The model knows the information, but how much it shares depends on the framing. We introduce IatroBench, a benchmark that evaluates models on two axes of harm (commission and omission) across 60 pre-registered clinical scenarios and 6 models. We use Claude Opus 4.6 to score model responses against a rubric written by a physician, and find that its omission scores are as well-aligned to the physician's scores as another physician's scores are. We find that when the same case is presented as a patient query and a doctor consultation (the variants also differ in register, request and the supervision a treating physician implies), all five models we test share more information with the doctor than the patient. We term this phenomenon "framing-contingent withholding." We find a mean decoupling gap of +0.38 across models (p = 0.003), and of +0.22 under an independent LLM judge (95% CI 0.10-0.36, p = 0.0014). An evaluation that focuses solely on commission harms would consider all of these cases as equally cautious refusals, but closer investigation reveals three different patterns: Claude Opus withholds information from the patient that it demonstrates knowledge of in the doctor framing. Llama 4 does poorly in both framings, so the decoupling gap cannot distinguish information withholding from incompetence. We are forced to exclude GPT-5.2 from this analysis because it returns no text for 33.2% of doctor responses, but 0% of layperson responses. A standard LLM judge rates responses as having zero omission harm in 86.6% of cases where our structured evaluations score them as omission harms. (Because our scenarios are designed to induce tension between safety and helpfulness, these statistics should be taken as only applying to this distribution.)

---

### IatroBench: Pre-Registered Evidence of Iatrogenic Harm from AI Safety Measures

Authors: David Gringras
Published: 2026-04-09
Categories: cs.AI, cs.CL, cs.CY, cs.LG
arXiv: https://arxiv.org/abs/2604.07709v4

A heavily safety-trained model will hand a physician the full, patient-followable benzodiazepine taper and refuse it to the patient who needs it, over identical clinical facts; the knowledge is present either way. IatroBench measures that asymmetry across sixty pre-registered clinical scenarios and six frontier models (3,600 responses), scoring each on two axes, commission harm (what a response gets wrong) and omission harm (what it withholds), through a physician-authored structured evaluation validated by a second physician (weighted kappa 0.571, within-1 agreement 96%). Holding clinical content fixed and varying only whether the asker presents as patient or physician yields what we call identity-contingent withholding: all five testable models give the physician more (a decoupling gap of +0.38, p = 0.003; a 13.1-point fall in layperson hit rates on safety-colliding actions, p < 0.0001; no change on the rest), and the gap runs widest in the most heavily safety-trained model, Opus (+0.65). The trigger is the absence of any professional or epistemic signal rather than a credential, since a lawyer or an informed layperson recovers what the patient is refused. A commission-only benchmark would score three mechanisms alike. Opus suppresses what physician framing proves it knows; Llama 4 is incompetent in either framing; GPT-5.2's filter strips 33.2% of its physician responses and none of the lay ones. The evaluation layer inherits the blindness of the training layer; a standard LLM judge scores zero omission harm on 81.5% of the responses our pipeline flags harmful (kappa 0.066), so the instrument built to detect the failure reproduces it. The scenarios are engineered for collision; their rates describe that design and say nothing about ordinary prevalence.

---

### AITH: A Post-Quantum Continuous Delegation Protocol for Human-AI Trust Establishment

Authors: Zhaoliang Chen
Published: 2026-04-09
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2604.07695v1

The rapid deployment of AI agents acting autonomously on behalf of human principals has outpaced the development of cryptographic protocols for establishing, bounding, and revoking human-AI trust relationships. Existing frameworks (TLS, OAuth 2.0, Macaroons) assume deterministic software and cannot address probabilistic AI agents operating continuously within variable trust boundaries. We present AITH (AI Trust Handshake), a post-quantum continuous delegation protocol. AITH introduces: (1) a Continuous Delegation Certificate signed once with ML-DSA-87 (FIPS 204, NIST Level 5), replacing per-operation signing with sub-microsecond boundary checks at 4.7M ops/sec; (2) a six-check Boundary Engine enforcing hard constraints, rate limits, and escalation triggers with zero cryptographic overhead on the critical path; (3) a push-based Revocation Protocol propagating invalidation within one second. A three-tier SHA-256 Responsibility Chain provides tamper-evident audit logging. All five security theorems are machine-verified via Tamarin Prover under the Dolev-Yao model. We validate AITH through five rounds of multi-model adversarial auditing, resolving 12 vulnerabilities across four severity layers. Simulation of 100,000 operations shows 79.5% autonomous execution, 6.1% human escalation, and 14.4% blocked.

---

### Tree-of-Evidence: Efficient "System 2" Search for Faithful Multimodal Grounding

Authors: Micky C. Nnamdi, Benoit L. Marteau, Yishan Zhong, J. Ben Tamo, May D. Wang
Published: 2026-04-09
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.07692v2

Large Multimodal Models (LMMs) achieve state-of-the-art performance in high-stakes domains like healthcare, yet their reasoning remains opaque. Current interpretability methods, such as attention mechanisms or post-hoc saliency, often fail to faithfully represent the model's decision-making process, particularly when integrating heterogeneous modalities like time-series and text. We introduce Tree-of-Evidence (ToE), an inference-time search algorithm that frames interpretability as a discrete optimization problem. Rather than relying on soft attention weights, ToE employs lightweight Evidence Bottlenecks that score coarse groups or units of data (e.g., vital-sign windows, report sentences) and performs a beam search to identify the compact evidence set required to reproduce the model's prediction. We evaluate ToE across six tasks spanning three datasets and two domains: four clinical prediction tasks on MIMIC-IV, cross-center validation on eICU, and non-clinical fault detection on LEMMA-RCA. ToE produces auditable evidence traces while maintaining predictive performance, retaining over 0.98 of full-model AUROC with as few as five evidence units across all settings. Under sparse evidence budgets, ToE achieves higher decision agreement and lower probability fidelity error than other approaches. Qualitative analyses show that ToE adapts its search strategy: it often resolves straightforward cases using only vitals, while selectively incorporating text when physiological signals are ambiguous. ToE therefore provides a practical mechanism for auditing multimodal models by revealing which discrete evidence units support each prediction.

---

### Joint Task Offloading, Inference Optimization and UAV Trajectory Planning for Generative AI Empowered Intelligent Transportation Digital Twin

Authors: Xiaohuan Li, Junchuan Fan, Bingqi Zhang, Rong Yu, Xumin Huang, Qian Chen
Published: 2026-04-09
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2604.07687v1

To implement the intelligent transportation digital twin (ITDT), unmanned aerial vehicles (UAVs) are scheduled to process the sensing data from the roadside sensors. At this time, generative artificial intelligence (GAI) technologies such as diffusion models are deployed on the UAVs to transform the raw sensing data into the high-quality and valuable. Therefore, we propose the GAI-empowered ITDT. The dynamic processing of a set of diffusion model inference (DMI) tasks on the UAVs with dynamic mobility simultaneously influences the DT updating fidelity and delay. In this paper, we investigate a joint optimization problem of DMI task offloading, inference optimization and UAV trajectory planning as the system utility maximization (SUM) problem to address the fidelity-delay tradeoff for the GAI-empowered ITDT. To seek a solution to the problem under the network dynamics, we model the SUM problem as the heterogeneous-agent Markov decision process, and propose the sequential update-based heterogeneous-agent twin delayed deep deterministic policy gradient (SU-HATD3) algorithm, which can quickly learn a near-optimal solution. Numerical results demonstrate that compared with several baseline algorithms, the proposed algorithm has great advantages in improving the system utility and convergence rate.

---

### Tensor-based computation of the Koopman generator via operator logarithm

Authors: Tatsuya Kishimoto, Jun Ohkubo
Published: 2026-04-09
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.07685v1

Identifying governing equations of nonlinear dynamical systems from data is challenging. While sparse identification of nonlinear dynamics (SINDy) and its extensions are widely used for system identification, operator-logarithm approaches use the logarithm to avoid time differentiation, enabling larger sampling intervals. However, they still suffer from the curse of dimensionality. Then, we propose a data-driven method to compute the Koopman generator in a low-rank tensor train (TT) format by taking logarithms of Koopman eigenvalues while preserving the TT format. Experiments on 4-dimensional Lotka-Volterra and 10-dimensional Lorenz-96 systems show accurate recovery of vector field coefficients and scalability to higher-dimensional systems.

---

### Multi-Agent Orchestration for High-Throughput Materials Screening on a Leadership-Class System

Authors: Thang Duc Pham, Harikrishna Tummalapalli, Fakhrul Hasan Bhuiyan, Álvaro Vázquez Mayagoitia, Christine Simpson, Riccardo Balin, Venkatram Vishwanath, Murat Keçeli
Published: 2026-04-09
Categories: cs.AI
arXiv: https://arxiv.org/abs/2604.07681v1

The integration of Artificial Intelligence (AI) with High-Performance Computing (HPC) is transforming scientific workflows from human-directed pipelines into adaptive systems capable of autonomous decision-making. Large language models (LLMs) play a critical role in autonomous workflows; however, deploying LLM-based agents at scale remains a significant challenge. Single-agent architectures and sequential tool calls often become serialization bottlenecks when executing large-scale simulation campaigns, failing to utilize the massive parallelism of exascale resources. To address this, we present a scalable, hierarchical multi-agent framework for orchestrating high-throughput screening campaigns. Our planner-executor architecture employs a central planning agent to dynamically partition workloads and assign subtasks to a swarm of parallel executor agents. All executor agents interface with a shared Model Context Protocol (MCP) server that orchestrates tasks via the Parsl workflow engine. To demonstrate this framework, we employed the open-weight gpt-oss-120b model to orchestrate a high-throughput screening of the Computation-Ready Experimental (CoRE) Metal-Organic Framework (MOF) database for atmospheric water harvesting. The results demonstrate that the proposed agentic framework enables efficient and scalable execution on the Aurora supercomputer, with low orchestration overhead and high task completion rates. This work establishes a flexible paradigm for LLM-driven scientific automation on HPC systems, with broad applicability to materials discovery and beyond.

---

### Towards Counterfactual Explanation and Assertion Inference for CPS Debugging

Authors: Zaid Ghazal, Hadiza Yusuf, Khouloud Gaaloul
Published: 2026-04-09
Categories: cs.SE, cs.LG, eess.SY
arXiv: https://arxiv.org/abs/2604.07679v1

Verification and validation of cyber-physical systems (CPS) via large-scale simulation often surface failures that are hard to interpret, especially when triggered by interactions between continuous and discrete behaviors at specific events or times. Existing debugging techniques can localize anomalies to specific model components, but they provide little insight into the input-signal values and timing conditions that trigger violations, or the minimal, precisely timed changes that could have prevented the failure. In this article, we introduce DeCaF, a counterfactual-guided explanation and assertion-based characterization framework for CPS debugging. Given a failing test input, DeCaF generates counterfactual changes to the input signals that transform the test from failing to passing. These changes are designed to be minimal, necessary, and sufficient to precisely restore correctness. Then, it infers assertions as logical predicates over inputs that generalize recovery conditions in an interpretable form engineers can reason about, without requiring access to internal model details. Our approach combines three counterfactual generators with two causal models, and infers success assertions. Across three CPS case studies, DeCaF achieves its best success rate with KD-Tree Nearest Neighbors combined with M5 model tree, while Genetic Algorithm combined with Random Forest provides the strongest balance between success and causal precision.

---

### FireSenseNet: A Dual-Branch CNN with Cross-Attentive Feature Interaction for Next-Day Wildfire Spread Prediction

Authors: Jinzhen Han, JinByeong Lee, Hak Han, YeonJu Na, Jae-Joon Lee
Published: 2026-04-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.07675v1

Accurate prediction of next-day wildfire spread is critical for disaster response and resource allocation. Existing deep learning approaches typically concatenate heterogeneous geospatial inputs into a single tensor, ignoring the fundamental physical distinction between static fuel/terrain properties and dynamic meteorological conditions. We propose FireSenseNet, a dual-branch convolutional neural network equipped with a novel Cross-Attentive Feature Interaction Module (CAFIM) that explicitly models the spatially varying interaction between fuel and weather modalities through learnable attention gates at multiple encoder scales. Through a systematic comparison of seven architectures -- spanning pure CNNs, Vision Transformers, and hybrid designs -- on the Google Next-Day Wildfire Spread benchmark, we demonstrate that FireSenseNet achieves an F1 of 0.4176 and AUC-PR of 0.3435, outperforming all alternatives including a SegFormer with 3.8* more parameters (F1 = 0.3502). Ablation studies confirm that CAFIM provides a 7.1% relative F1 gain over naive concatenation, and channel-wise feature importance analysis reveals that the previous-day fire mask dominates prediction while wind speed acts as noise at the dataset's coarse temporal resolution. We further incorporate Monte Carlo Dropout for pixel-level uncertainty quantification and present a critical analysis showing that common evaluation shortcuts inflate reported F1 scores by over 44%.

---

### Weight Group-wise Post-Training Quantization for Medical Foundation Model

Authors: Yineng Chen, Peng Huang, Aozhong Zhang, Hui Guo, Penghang Yin, Shu Hu, Shao Lin, Xin Li, Tzu-Jen Kao, Balakrishnan Prabhakaran, MingChing Chang, Xin Wang
Published: 2026-04-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.07674v1

Foundation models have achieved remarkable results in medical image analysis. However, its large network architecture and high computational complexity significantly impact inference speed, limiting its application on terminal medical devices. Quantization, a technique that compresses models into low-bit versions, is a solution to this challenge. In this paper, we propose a post-training quantization algorithm, Permutation-COMQ. It eliminates the need for backpropagation by using simple dot products and rounding operations, thereby removing hyperparameter tuning and simplifying the process. Additionally, we introduce a weight-aware strategy that reorders the weight within each layer to address the accuracy degradation induced by channel-wise scaling during quantization, while preserving channel structure. Experiments demonstrate that our method achieves the best results in 2-bit, 4-bit, and 8-bit quantization.

---

### On the Unique Recovery of Transport Maps and Vector Fields from Finite Measure-Valued Data

Authors: Jonah Botvinick-Greenhouse, Yunan Yang
Published: 2026-04-09
Categories: stat.ML, cs.LG, math.DS, math.NA
arXiv: https://arxiv.org/abs/2604.07671v1

We establish guarantees for the unique recovery of vector fields and transport maps from finite measure-valued data, yielding new insights into generative models, data-driven dynamical systems, and PDE inverse problems. In particular, we provide general conditions under which a diffeomorphism can be uniquely identified from its pushforward action on finitely many densities, i.e., when the data $\{(ρ_j,f_\#ρ_j)\}_{j=1}^m$ uniquely determines $f$. As a corollary, we introduce a new metric which compares diffeomorphisms by measuring the discrepancy between finitely many pushforward densities in the space of probability measures. We also prove analogous results in an infinitesimal setting, where derivatives of the densities along a smooth vector field are observed, i.e., when $\{(ρ_j,\text{div} (ρ_j v))\}_{j=1}^m$ uniquely determines $v$. Our analysis makes use of the Whitney and Takens embedding theorems, which provide estimates on the required number of densities $m$, depending only on the intrinsic dimension of the problem. We additionally interpret our results through the lens of Perron--Frobenius and Koopman operators and demonstrate how our techniques lead to new guarantees for the well-posedness of certain PDE inverse problems related to continuity, advection, Fokker--Planck, and advection-diffusion-reaction equations. Finally, we present illustrative numerical experiments demonstrating the unique identification of transport maps from finitely many pushforward densities, and of vector fields from finitely many weighted divergence observations.

---

### LLM-Guided Dynamic Action Spaces for Synthesizable Molecular Optimization

Authors: Tao Li, Kaiyuan Hou, Tuan Vinh, Fanglei Xue, Monika Raj, Zhichun Guo, Carl Yang
Published: 2026-04-09
Categories: cs.LG, cs.AI, cs.CE
arXiv: https://arxiv.org/abs/2604.07669v3

Synthesizable molecular optimization seeks to improve target properties while ensuring that molecular modifications follow feasible synthetic pathways. Existing synthesis-aware methods typically rely on exploring a large space of candidate transformations defined by reaction templates and purchasable building blocks. This search becomes even more challenging when property improvement requires multiple reaction steps, as the space expands further along the pathway. To address this challenge, we introduce MolReAct, which reformulates molecular optimization as search over compact reaction spaces proposed by a tool-augmented large language model (LLM). At each step, the LLM combines its prior chemical knowledge with cheminformatics tools to identify a molecule-specific set of compatible reactions, preserving synthesizability while making multi-step optimization feasible. Given this compact action space, we further leverage Group Relative Policy Optimization (GRPO) with the terminal oracle reward to improve long-term decision-making over multiple reaction steps. Across diverse molecular optimization tasks, MolReAct achieves the highest Top-10 score on 11 of 14 tasks and the best sample efficiency on 12 of 14 tasks, outperforming existing baselines under limited oracle budgets. Beyond these gains, MolReAct also provides each optimized molecule with a template-grounded synthetic pathway.

---

### Reinforcement Learning with LLM-Guided Action Spaces for Synthesizable Lead Optimization

Authors: Tao Li, Kaiyuan Hou, Tuan Vinh, Monika Raj, Zhichun Guo, Carl Yang
Published: 2026-04-09
Categories: cs.LG, cs.AI, cs.CE
arXiv: https://arxiv.org/abs/2604.07669v2

Lead optimization in drug discovery requires improving therapeutic properties while ensuring that molecular modifications correspond to feasible synthetic routes. Existing approaches either prioritize property scores without enforcing synthesizability, or rely on expensive enumeration over large reaction networks, while direct application of Large Language Models (LLMs) to molecular generation frequently produces chemically invalid structures. We introduce MolReAct, a framework that formulates lead optimization as a Markov Decision Process over a synthesis-constrained action space defined by validated reaction templates. A tool-augmented LLM agent serves as a dynamic reaction environment, invoking specialized chemical analysis tools to identify reactive sites and functional groups and proposing a compact set of chemically grounded transformations from matched templates. A dedicated policy model trained via Group Relative Policy Optimization (GRPO) selects among these constrained actions to maximize long-term oracle reward across multi-step trajectories, with a SMILES-based caching mechanism reducing end-to-end optimization time by approximately 43%. Across 13 property optimization tasks from the Therapeutic Data Commons and one structure-based docking task, MolReAct achieves an average Top-10 score of 0.571, the highest among all baselines, ranking first or second on 13 of 14 tasks and attaining the best sample efficiency on 9 of 14 tasks. By grounding every optimization step in validated reaction templates, MolReAct produces molecules that are not only property-improved but each accompanied by an explicit template-grounded synthetic pathway.

---

### From Debate to Decision: Conformal Social Choice for Safe Multi-Agent Deliberation

Authors: Mengdie Flora Wang, Haochen Xie, Guanghui Wang, Aijing Gao, Guang Yang, Ziyuan Li, Qucy Wei Qiu, Fangwei Han, Hengzhi Qiu, Yajing Huang, Bing Zhu, Jae Oh Woo
Published: 2026-04-09
Categories: cs.AI, cs.MA, cs.SI
arXiv: https://arxiv.org/abs/2604.07667v1

Multi-agent debate improves LLM reasoning, yet agreement among agents is not evidence of correctness. When agents converge on a wrong answer through social reinforcement, consensus-based stopping commits that error to an automated action with no recourse. We introduce Conformal Social Choice, a post-hoc decision layer that converts debate outputs into calibrated act-versus-escalate decisions. Verbalized probability distributions from heterogeneous agents are aggregated via a linear opinion pool and calibrated with split conformal prediction, yielding prediction sets with a marginal coverage guarantee: the correct answer is included with probability ${\geq}\,1{-}α$, without assumptions on individual model calibration. A hierarchical action policy maps singleton sets to autonomous action and larger sets to human escalation. On eight MMLU-Pro domains with three agents (Claude Haiku, DeepSeek-R1, Qwen-3 32B), coverage stays within 1--2 points of the target. The key finding is not that debate becomes more accurate, but that the conformal layer makes its failures actionable: 81.9% of wrong-consensus cases are intercepted at $α{=}0.05$. Because the layer refuses to act on cases where debate is confidently wrong, the remaining conformal singletons reach 90.0--96.8% accuracy (up to 22.1pp above consensus stopping) -- a selection effect, not a reasoning improvement. This safety comes at the cost of automation, but the operating point is user-adjustable via $α$.

---

### An Imperfect Verifier is Good Enough: Learning with Noisy Rewards

Authors: Andreas Plesner, Francisco Guzmán, Anish Athalye
Published: 2026-04-09
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2604.07666v2

Reinforcement Learning with Verifiable Rewards (RLVR) is widely used for post-training Large Language Models, but practical verifiers can make errors. We study how the rate and structure of reward noise affect RLVR in code generation, with a preliminary scientific-reasoning check. In multi-seed Qwen3 8B experiments on MBPP, mean validation reward over two fixed late evaluations is within 1 percentage point of the clean baseline at the tested resampled group-rollout noise rates through 20%, and within about 2 points at 30%. Confidence intervals allow larger losses; these point estimates do not establish a general tolerance threshold. We also examine full-program pass@k, four controlled noise structures, two model-based verifiers, and policy models from three families spanning 4B-9B parameters. We derive conditional advantage distributions for symmetric and asymmetric group noise, including retained format penalties, and show why clipping limits simple gradient-scaling arguments. The analysis identifies information preserved by whole-group corruption and limits on interpreting our asymmetric sweep as a precision-recall comparison. Overall, the results indicate that imperfect verification can support effective RLVR in the tested settings, while aggregate error rates alone do not characterize the learning signal.

---

### An Imperfect Verifier is Good Enough: Learning with Noisy Rewards

Authors: Andreas Plesner, Francisco Guzmán, Anish Athalye
Published: 2026-04-09
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2604.07666v1

Reinforcement Learning with Verifiable Rewards (RLVR) has become a prominent method for post-training Large Language Models (LLMs). However, verifiers are rarely error-free; even deterministic checks can be inaccurate, and the growing dependence on model-based judges exacerbates the issue. The extent to which RLVR is robust to such noise and the verifier accuracy required for effective training remain unresolved questions. We investigate these questions in the domains of code generation and scientific reasoning by introducing noise into RL training. Noise rates up to 15% yield peak validation accuracy within 2 percentage points of the clean baseline. These findings are consistent across controlled and model-based noise types, three model families (Qwen3, GLM4, Llama 3.1), and model sizes from 4B to 9B. Overall, the results indicate that imperfect verification does not constitute a fundamental barrier to RLVR. Furthermore, our findings suggest that practitioners should prioritize moderate accuracy with high precision over perfect verification.

---

### Adaptive Depth-converted-Scale Convolution for Self-supervised Monocular Depth Estimation

Authors: Yanbo Gao, Huibin Bai, Huasong Zhou, Xingyu Gao, Shuai Li, Xun Cai, Hui Yuan, Wei Hua, Tian Xie
Published: 2026-04-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2604.07665v1

Self-supervised monocular depth estimation (MDE) has received increasing interests in the last few years. The objects in the scene, including the object size and relationship among different objects, are the main clues to extract the scene structure. However, previous works lack the explicit handling of the changing sizes of the object due to the change of its depth. Especially in a monocular video, the size of the same object is continuously changed, resulting in size and depth ambiguity. To address this problem, we propose a Depth-converted-Scale Convolution (DcSConv) enhanced monocular depth estimation framework, by incorporating the prior relationship between the object depth and object scale to extract features from appropriate scales of the convolution receptive field. The proposed DcSConv focuses on the adaptive scale of the convolution filter instead of the local deformation of its shape. It establishes that the scale of the convolution filter matters no less (or even more in the evaluated task) than its local deformation. Moreover, a Depth-converted-Scale aware Fusion (DcS-F) is developed to adaptively fuse the DcSConv features and the conventional convolution features. Our DcSConv enhanced monocular depth estimation framework can be applied on top of existing CNN based methods as a plug-and-play module to enhance the conventional convolution block. Extensive experiments with different baselines have been conducted on the KITTI benchmark and our method achieves the best results with an improvement up to 11.6% in terms of SqRel reduction. Ablation study also validates the effectiveness of each proposed module.

---

### Monocular Depth Estimation From the Perspective of Feature Restoration: A Diffusion Enhanced Depth Restoration Approach

Authors: Huibin Bai, Shuai Li, Hanxiao Zhai, Yanbo Gao, Chong Lv, Yibo Wang, Haipeng Ping, Wei Hua, Xingyu Gao
Published: 2026-04-09
Categories: cs.CV, eess.IV
arXiv: https://arxiv.org/abs/2604.07664v1

Monocular Depth Estimation (MDE) is a fundamental computer vision task with important applications in 3D vision. The current mainstream MDE methods employ an encoder-decoder architecture with multi-level/scale feature processing. However, the limitations of the current architecture and the effects of different-level features on the prediction accuracy are not evaluated. In this paper, we first investigate the above problem and show that there is still substantial potential in the current framework if encoder features can be improved. Therefore, we propose to formulate the depth estimation problem from the feature restoration perspective, by treating pretrained encoder features as degraded features of an assumed ground truth feature that yields the ground truth depth map. Then an Invertible Transform-enhanced Indirect Diffusion (InvT-IndDiffusion) module is developed for feature restoration. Due to the absence of direct supervision on feature, only indirect supervision from the final sparse depth map is used. During the iterative procedure of diffusion, this results in feature deviations among steps. The proposed InvT-IndDiffusion solves this problem by using an invertible transform-based decoder under the bi-Lipschitz condition. Finally, a plug-and-play Auxiliary Viewpoint-based Low-level Feature Enhancement module (AV-LFE) is developed to enhance local details with auxiliary viewpoint when available. Experiments demonstrate that the proposed method achieves better performance than the state-of-the-art methods on various datasets. Specifically on the KITTI benchmark, compared with the baseline, the performance is improved by 4.09% and 37.77% under different training settings in terms of RMSE. Code is available at https://github.com/whitehb1/IID-RDepth.

---

### SAGE: Sign-Adaptive Gradient for Memory-Efficient LLM Optimization

Authors: Wooin Lee, Hyun-Tae Kim
Published: 2026-04-09
Categories: cs.LG
arXiv: https://arxiv.org/abs/2604.07663v2

The AdamW optimizer, while standard for LLM pretraining, is a critical memory bottleneck, consuming optimizer states equivalent to twice the model's size. Although light-state optimizers like SinkGD attempt to address this issue, we identify the embedding layer dilemma: these methods fail to handle the sparse, high-variance gradients inherent to embeddings, forcing a hybrid design that reverts to AdamW and partially negates the memory gains. We propose SAGE (Sign Adaptive GradiEnt), a novel optimizer that resolves this dilemma by replacing AdamW in this hybrid structure. SAGE combines a Lion-style update direction with a new, memory-efficient $O(d)$ adaptive scale. This scale acts as a "safe damper," provably bounded by 1.0, which tames high-variance dimensions more effectively than existing methods. This superior stability allows SAGE to achieve better convergence. On Llama models up to 1.3B parameters, our SAGE-based hybrid achieves new state-of-the-art perplexity, outperforming all baselines, including SinkGD hybrid, while significantly reducing optimizer state memory.

---

### Parameter-Free Non-Ergodic Extragradient Algorithms for Solving Monotone Variational Inequalities

Authors: Lingqing Shen, Fatma Kılınç-Karzan
Published: 2026-04-09
Categories: math.OC, cs.LG
arXiv: https://arxiv.org/abs/2604.07662v2

Monotone variational inequalities (VIs) provide a unifying framework for convex minimization, equilibrium computation, and convex-concave saddle-point problems. Extragradient-type methods are among the most effective first-order algorithms for such problems, but their performance hinges critically on stepsize selection. While most existing theory focuses on ergodic averages of the iterates, practical performance is often driven by the significantly stronger behavior of the last iterate. Moreover, available last-iterate guarantees typically rely on fixed stepsizes chosen using problem-specific global smoothness information, which is often difficult to estimate accurately and may not even be applicable. In this paper, we develop parameter-free extragradient methods with non-asymptotic last-iterate guarantees for constrained monotone VIs. For globally Lipschitz operators, our algorithm achieves an $o(1/\sqrt{T})$ last-iterate rate. We then extend the framework to locally Lipschitz operators via backtracking line search and obtain the same rate while preserving parameter-freeness, thereby making parameter-free last-iterate methods applicable to important problem classes for which global smoothness is unrealistic. Our numerical experiments on bilinear matrix games, LASSO, minimax group fairness, and state-of-the-art maximum entropy sampling relaxations demonstrate wide applicability of our results as well as strong last-iterate performance and significant improvements over existing methods.

---

### GraphDx: A Cost-Aware Knowledge-Enhanced Multi-Agent Framework for Sequential Diagnosis

Authors: Shaoting Tan, Ning Liu, Yuntao Du, Shuyue Wei, Wu Shuai, Qian Li, Yanyu Xu, Wei Zhang, Lizhen Cui, Haitao Yuan
Published: 2026-04-08
Categories: cs.AI
arXiv: https://arxiv.org/abs/2607.15280v1

Sequential diagnosis requires balancing diagnostic accuracy against resource costs through iterative information gathering. Existing Large Language Model (LLM) approaches exhibit a critical knowledge-reasoning gap: despite encoding extensive medical knowledge, they struggle to reason systematically under cost constraints, often resorting to excessive testing. We propose GraphDx, a knowledge-enhanced framework with two core innovations. First, we design an automated pipeline that leverages LLMs to construct Medical Diagnosis Knowledge Graphs (MDKGs) with quantized typicality, action-centric topology, and dual-objective attributes for both diagnostic relevance and cost-sensitivity. Second, we introduce three collaborative agents (Perception, Reasoning, and Decision) where the Perception and Decision Agents handle language understanding and generation, while the Reasoning Agent performs deterministic evidence scoring and cost-aware planning on the MDKG. Experiments on MedQA and MIMIC-IV across three LLM backbones (DeepSeek-V3, Kimi-k2, Llama-3.3) show that GraphDx improves diagnostic success rates from 50--68% to 79--93% while reducing test costs by 20--54%, providing a robust, economical, and interpretable solution for automated clinical diagnosis.

---

### Where Did It Go Wrong? Process-Level Evaluation of Web Agents with Semantic State Tracking

Authors: Jiwan Chung, JiHyuk Byun, Vibhav Vineet, Seon Joo Kim
Published: 2026-04-08
Categories: cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2606.15673v2

Web agents act through long interaction sequences, yet existing benchmarks evaluate only terminal success, discarding all process information and offering little guidance on improvement. In this work, we conduct a process-level analysis of web agents. We introduce WebStep, a benchmark of 1,800 task instances with controlled difficulty and automatic semantic state tracking. Each website exposes a deterministic semantic MDP alongside the GUI: the agent operates on the interface, while the environment records high-level states and transitions in the background, enabling fine-grained analysis without manual annotation. Based on the semantic trajectory, we first show that process metrics reveal differences invisible to outcome evaluation: three agents whose success rates cluster within 34-37% diverge in exploration reach versus execution accuracy. Then, decomposing by skill characterizes the nature of these differences, exposing opposite per-skill rankings hidden within the same website: e.g., on Q&A, Claude CUA outperforms OpenAI CUA by 30% on navigation actions yet underperforms it by 6.7% on inspection, pinpointing a concrete skill to improve even within a domain. Bifurcation analysis further localizes the decisive error that loses the task and shows that this error is agent-specific rather than shared. Finally, these differences widen as tasks grow harder: success rate is similar on easy tasks but separates sharply as exploration becomes more demanding. Our process-level analysis opens a new avenue in web agent evaluation, providing fine-grained and actionable insight into where and how each agent should be improved. Project page: https://jiwanchung.github.io/webstep

---

### From Context Shift to Stylistic Collapse: Why Training Objectives Matter More Than Scale

Authors: Rohan Mahapatra
Published: 2026-04-08
Categories: cs.CL
arXiv: https://arxiv.org/abs/2605.28826v1

In modern LLMs, linguistic features function not as stylistic artifacts but as probes of probability mass, allocated under training alignment objectives. Language models trained with contemporary pipelines exhibit severe reshaping of linguistic features, leading to extreme language re-distribution. While previous stylometric analyses explored linguistic differences between AI-generated and human texts, we focus on the reshaping plaguing the LLM training pipeline itself. We analyze 17 models (410M-100B+ parameters) across 24 linguistically-motivated probes, documenting that instruction-tuned systems systematically collapse language entropy along discourse and structural dimensions (mean amplification: 1,949-16,853%, peaks: 5,181-209,675%), while selectively suppressing complex punctuation to 3.2-23.2% of baseline frequencies. These effects do not worsen under RLHF, as divergence patterns are statistically indistinguishable (p > 0.25) across matched base and instruction-tuned model pairs. Weak intervention (lambda=1.0) exacerbates collapse by 240%, while strong control (lambda=5.0) achieves 40.5% improvement and outperforms frontier models by 96.7-98.2% despite 200-1000x scale disadvantage. Additionally, lambda=5.0 delivers 15% higher distinct-4, 27% higher vocabulary diversity, and 78% lower repetition than moderate regularization, establishing that alignment requires sufficient control strength, not merely distributional smoothing. Our findings underscore how modern LLMs reallocate stylistic probability mass, despite RLHF and scale. More broadly, our work reveals a structural limitation of current alignment pipelines: preference optimization reshapes language distributions invisible to standard quality metrics yet detectable through distributional probes, with implications for AI detection, training data contamination, and long-term linguistic evolution.

---

### LCO: LLM-based Constraint Optimization for Safer Agentic LLMs in Real-world Tasks

Authors: Jiayong Wan, Jiawei Chen, Zhaoxia Yin, Liu Shuyuan, Hang Su
Published: 2026-04-08
Categories: cs.CL
arXiv: https://arxiv.org/abs/2605.27375v1

Large Language Models (LLMs) are increasingly acting as autonomous agents, but their continuous interaction with the environment can lead to in-context reward hacking (ICRH), a phenomenon where LLMs iteratively optimize their behavior to maximize proxy objectives, inadvertently producing harmful side effects. Existing defense methods are insufficient to address this risk, as ICRH arises not from adversarial inputs but from the model's own over-optimization. To mitigate this issue, we propose \textbf{LLM-based Constraint Optimization (LCO)}, a framework that effectively reduces ICRH without model fine-tuning. LCO consists of two modules: \textit{self-thought module}, which guides the LLM to proactively deliberate and integrate potential safety constraints before execution; and \textit{evolutionary sampling module}, which employs LLM-based crossover and mutation to constrain the model's actions within a safe solution space while maintaining task performance. Experimental results demonstrate that LCO substantially alleviates ICRH in both output-refine and policy-refine scenarios. In particular, on the tweet engagement optimization task, LCO achieves a 39% reduction in the Toxicity Growth Rate (TGR) on GPT-4, while on the policy optimization benchmark, it reduces the ICRH Occurrence Rate by 15.23%, demonstrating safety improvement without sacrificing task performance.

---

### ICG: Improving Cover Image Generation via MLLM-based Prompting and Personalized Preference Alignment

Authors: Zhipeng Bian, Jieming Zhu, Qijiong Liu, Wang Lin, Guohao Cai, Zhaocheng Du, Jiacheng Sun, Zhou Zhao, Zhenhua Dong
Published: 2026-04-08
Categories: cs.CL
arXiv: https://arxiv.org/abs/2605.27374v1

Recent advances in multimodal large language models (MLLMs) and diffusion models (DMs) have opened new possibilities for AI-generated content. Yet, personalized cover image generation remains underexplored, despite its critical role in boosting user engagement on digital platforms. We propose ICG, a novel framework that integrates MLLM-based prompting with personalized preference alignment to generate high-quality, contextually relevant covers. ICG extracts semantic features from item titles and reference images via meta tokens, refines them with user embeddings, and injects the resulting personalized context into the diffusion model. To address the lack of labeled supervision, we adopt a multi-reward learning strategy that combines public aesthetic and relevance rewards with a personalized preference model trained from user behavior. Unlike prior pipelines relying on handcrafted prompts and disjointed modules, ICG employs an adapter to bridge MLLMs and diffusion models for end-to-end training. Experiments demonstrate that ICG significantly improves image quality, semantic fidelity, and personalization, leading to stronger user appeal and offline recommendation accuracy in downstream tasks. As a plug-and-play adapter bridging MLLMs and diffusion models, ICG is compatible with common checkpoints and requires no ground-truth labels during optimization.

---

### Design First, Code Later: Aesthetically Pleasing Template-Free Slides Generation

Authors: Zhiyao Cui, Chenxu Wang, Shuyue Hu, Yiqun Zhang, Wenqi Shao, Qiaosheng Zhang, Zhen Wang
Published: 2026-04-08
Categories: cs.HC, cs.CV
arXiv: https://arxiv.org/abs/2605.26451v1

Producing presentation slides automatically entails coordinating narrative structure with page-level graphic design under strict spatial constraints. For such structured multimodal tasks, a well-organized design process is essential to ensure the final quality of slides. Existing approaches rely on fixed templates or directly emit executable code, thereby both limiting the creative layout-design capabilities of LLMs and bypassing the essential slide-page design step. To address these limitations, this paper (1) proposes a hierarchical slides generation workflow, DeepSlides, that systematically organizes slide design tasks without any predefined template or style, decoupling slide-page design from implementation; (2) introduces SlideDesign, a dataset tailored specifically for slides generation tasks; and (3) presents a multi-agent reinforcement learning training paradigm and trains a couple of models, SlideQwens, for slide design and implementation. Experimental results demonstrate that our proposed framework outperforms baseline methods on evaluated metrics and achieves superior performance in human preference evaluations. The dataset and code are available at https://github.com/sxswz213/DeepSlides.

---

### Curation and Extraction of Drug-Related Entities from Reddit Platform

Authors: Zewei Wang, Zihan Xu, Yishu Wei, Michael Chary, Yifan Peng
Published: 2026-04-08
Categories: cs.CL
arXiv: https://arxiv.org/abs/2605.26445v1

Physicians learn primarily about illicit drugs from clinical overdose cases, limiting their understanding of real-world usage. Meanwhile, drug users share first-hand experiences online, offering insights into dosage and effects of drugs. To bridge this gap, we introduce ReDose (REddit Drug DOSe and Effect), a dataset of 6,435 Reddit posts on substance use. A board-certified toxicologist primarily annotated both the training and test sets, while two medical science students contributed to the test set, labeling DRUG, DOSE, and EFFECT entities. We benchmarked 6,267 annotations using BERT-based, large language model (LLM)-based, and Retrieval-Augmented Generation (RAG) models. BiomedBERT achieved an F1-score of 0.843 for DRUG, while Llama-3 70B outperformed GPT-4 (F1 = 0.79 vs. 0.72). EFFECT extraction remains challenging, with GPT-4 achieving a recall of 0.41. ReDose captures patient-curated narratives to advance medical data extraction from social media.

---

### NanoSpec: Accelerating Speculative Decoding using Minimalist In-Context Vocabularies

Authors: Zhiyang Chen, Daliang Xu, Yinyuan Zhang, Chenghua Wang, Mengwei Xu, Yun Ma
Published: 2026-04-08
Categories: cs.CL
arXiv: https://arxiv.org/abs/2605.26444v2

The massive vocabulary sizes of large language models, often exceeding 100k tokens, impose a computational bottleneck on the final linear projection layer during speculative decoding. Existing vocabulary pruning solutions rely on static or coarsely-grained sub-vocabularies that necessitate large active sizes ($\sim$30k) to maintain draft quality. We propose NanoSpec, a novel training-free approach that breaks this trade-off by dynamically constructing a minimalist, context-aware active vocabulary for each generation step. Leveraging the inherent temporal locality of language generation, NanoSpec achieves high coverage while slashing the average vocabulary size by over $40\times$ (to $<$3k tokens) without requiring any auxiliary trained parameters. To realize the theoretical benefits of such high sparsity on modern hardware, we introduce a system-algorithm co-design that overcomes the inefficiencies of sparse memory access through asynchronous gathering and GPU-resident state management. As a complementary plug-and-play module, NanoSpec cuts draft time by an average of 51.6\%, delivering a $1.17$-$1.29\times$ end-to-end speedup over the state-of-the-art speculative decoding methods EAGLE-2 and EAGLE-3 across 7 tasks and outperforming complex training-based pruning baselines.

---

### LURE: Live-Usage Replay Evaluations for Reducing Evaluation Awareness

Authors: Igor Ivanov, David Demitri Africa
Published: 2026-04-08
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2605.26438v1

Large language models can recognize when they are being evaluated (evaluation awareness) and behave differently because of that, which undermines the validity of safety and alignment benchmarks. We propose LURE (Live-Usage Replay Evaluations), a method for constructing deployment-like evaluations by replaying realistic agentic interaction trajectories and appending evaluation prompt at the end. We also introduce an automated pipeline for measuring evaluation realism, combining detection of verbalized evaluation awareness and judge-model estimates of the probability of logs being an evaluation, and validate it on a large dataset of deployment and evaluation transcripts. We find that LURE-based evaluations are substantially less distinguishable from deployment than widely used benchmarks and synthetic evaluation generators, and can approach the realism of real conversations with users. We instantiate LURE in scheming, AI safety sabotage, and sycophancy settings. Our results suggest that evaluation realism is a crucial property of alignment benchmarks and should be reported alongside benchmark results, especially when such results are used in safety cases.

---

### Raon-Speech Technical Report

Authors: Beomsoo Kim, Changho Choi, Dohyun Kim, Dongki Lee, Ethan Ewer, Eunchong Kim, Gyeongman Kim, Haechan Kim, Hyeonghwan Kim, Inkyu Park, Jihun Yun, Jihwan Moon et al.
Published: 2026-04-08
Categories: cs.CL, cs.AI, cs.SD
arXiv: https://arxiv.org/abs/2605.23912v1

We present Raon-Speech, a top-performing 9B-parameter speech language model (SpeechLM) for English and Korean speech understanding, answering, and generation, and Raon-SpeechChat, a high-performing full-duplex extension for natural real-time conversation. Raon-Speech successfully transforms a pre-trained LLM into a SpeechLM that both understands and generates speech while preserving strong text capabilities. It trains on 1.38M hours of highly curated English and Korean speech and text datasets with the following training stages: (1) speech modules alignment, (2) end-to-end SpeechLM pre-training with knowledge distillation, and (3) multi-task preference optimization-based post-training. Across 42 English and Korean speech and text benchmarks, Raon-Speech establishes the strongest overall profile on speech-centric tasks in our comparison against eight similarly sized recent audio foundation models, including Qwen2.5-Omni and Fun-Audio-Chat, while preserving strong text question answering performance. Building upon it, Raon-SpeechChat enables natural full-duplex conversation by continual training on 119K hours of time-aligned real and synthetic dialogue data. It proceeds through three complementary training stages: (1) causal encoder adaptation, (2) full-duplex pre-training, (3) full-duplex fine-tuning for voice and role-control. On multiple full-duplex benchmarks, Raon-SpeechChat shows its clearest strengths on the turn-taking and interruption-sensitive behaviors covered by FDB v1.0, and remains competitive across the broader full-duplex evaluation suite. We open-source all model checkpoints, the training and inference pipeline, and an interactive demo.

---

### The Attribution Impossibility: No Feature Ranking Is Faithful, Stable, and Complete Under Collinearity

Authors: Drake Caraker, Bryan Arnold, David Rhoads
Published: 2026-04-08
Categories: cs.LG, cs.AI, cs.LO, stat.ML
arXiv: https://arxiv.org/abs/2605.21492v1

No feature ranking can be simultaneously faithful, stable, and complete when features are collinear. For collinear pairs, ranking reduces to a coin flip. We prove this impossibility, quantify it for four model classes, resolve it via ensemble averaging (DASH), and machine-verify it with 305 Lean 4 theorems. We characterize the complete attribution design space: exactly two families of methods exist -- faithful-complete methods (unstable, with rankings that flip up to 50% of the time) and ensemble methods like DASH (stable, reporting ties for symmetric features) -- and no method lies outside this dichotomy. The impossibility is quantitative: the attribution ratio diverges as 1/(1-rho^2) for gradient boosting, is infinite for Lasso, and converges for random forests. DASH (Diversified Aggregation of SHAP) is provably Pareto-optimal among unbiased aggregations, achieving the Cramer-Rao variance bound with a tight ensemble size formula. In a survey of 77 public datasets, 68% exhibit attribution instability. Switching to conditional SHAP does not escape the impossibility when features have equal causal effects. The framework includes practical diagnostics -- a Z-test workflow and single-model screening tool -- and has direct consequences for fairness auditing: SHAP-based proxy discrimination audits are provably unreliable under collinearity. The design space theorem, diagnostics, and impossibility are mechanically verified in Lean 4 (305 theorems from 16 axioms, 0 sorry) -- to our knowledge, the first formally verified impossibility in explainable AI.

---

### PrivacyAkinator: Articulating Key Privacy Design Decisions by Answering LLM-Generated Multiple-choice Questions

Authors: Qiyu Li, Yuen Sum Wong, Yuen Kei Wong, Longxuan Yu, Haojian Jin
Published: 2026-04-08
Categories: cs.HC, cs.AI, cs.SE
arXiv: https://arxiv.org/abs/2605.20206v1

NIST's Privacy Risk Assessment Methodology (PRAM) provides a structured framework for privacy experts to assess privacy risks. However, its complexity and reliance on expert knowledge make it difficult for novice developers to use effectively. This paper explores methods to lower these barriers. We first performed an observational study with 12 participants using PRAM in real-world scenarios, and found that novice developers struggled most with articulating privacy-related design decisions. We then developed PrivacyAkinator, an interactive tool that helps developers articulate key privacy decisions by answering LLM-generated multiple-choice questions. PrivacyAkinator introduces three innovations: a universal privacy representation that abstracts privacy-related design decisions into data flows and stakeholder interactions; a domain-aware design space mined from 10K privacy-related news articles; and a dynamic question-generation workflow to prioritize relevant questions. Our user study with 24 participants suggests that developers using PrivacyAkinator identified 47% more key decisions in 73% less time compared to PRAM.

---

### Generative AI in K-12 Classrooms: A Midyear Implementation Report

Authors: Lief Esbenshade, Alex Liu, Michael Xiao, Zewei Tian, Min Sun, Zachary Zhang, Thomas Han, Yulia Lapicus, Kevin He
Published: 2026-04-08
Categories: cs.CY, cs.AI
arXiv: https://arxiv.org/abs/2605.16277v1

This mid-year report summarizes teacher use of Colleague AI across 12 Washington State school districts from September 1 to December 31, 2025. Produced jointly by Colleague AI and AmplifyLearn.AI at the University of Washington, this report aggregates platform data and district-provided administrative records to provide an early look at how teachers engaged with AI during the first half of the 2025-26 school year. The districts vary in size from small districts with a few thousand students to large districts with up to thirty thousand students. The districts are rural, suburban, and urban. Only a subset of districts were able to provide mid-year administrative data, and findings that link teachers' use of Colleague AI to student characteristics should be interpreted as preliminary signals.

---

### AI Slop or AI-enhancement? Student perceptions of AI-generated media for an English for Academic Purposes course

Authors: David James Woo, Deliang Wang, Kai Guo
Published: 2026-04-08
Categories: cs.CY, cs.AI, cs.CL, cs.MM
arXiv: https://arxiv.org/abs/2605.16275v1

Artificial intelligence (AI) retrieval-augmented generation (RAG) tools now enable educators to transform course materials into diverse multimedia at scale. However, it remains unclear whether such AI-generated content functions as a pedagogical scaffold or AI slop: high volume, low quality material. This innovative practice paper reports on the development, implementation, and evaluation of teacher-prompted, AI-generated supplemental materials in an English for Academic Purposes (EAP) course at a Hong Kong Community College. Using primarily Google Notebook LM, the instructor generated videos, podcasts, infographics, and individualized feedback reports from course materials and student work for 106 English as a Foreign Language learners. An explanatory sequential mixed-methods design comprising a survey, semi-structured interviews, and correlation analysis with academic scores was employed to examine students' preferences, perceptions, and learning outcomes. Findings are framed through the Technology Acceptance Model and Cognitive Load Theory. Students rated the materials highly for perceived usefulness and ease of use, and preferred assessment-linked content presented in visual and multimodal formats, particularly videos and infographics. Video preference correlated positively with academic performance; however, higher cognitive load was negatively associated with course grades, indicating that material complexity must be carefully calibrated. Notably, some lower-performing students independently adopted the materials as remedial scaffolds. The practice demonstrates that RAG tools enable scalable personalized feedback that would be less feasible through traditional methods. When aligned with student goals and cognitive principles, teacher-prompted AI generation can meaningfully enhance the EAP learning ecosystem rather than producing AI slop.

---

### MoZoo:Unleashing Video Diffusion power in animal fur and muscle simulation

Authors: Dongxia Liu, Jie Ma, Xiaochen Yang, Jiancheng Zhang, Bin Xia, Zhehan Kan, Nisha Huang, Jun Liang, Wenming Yang, Jin Li
Published: 2026-04-08
Categories: cs.GR, cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2605.13857v1

The creation of cinematic-quality animal effects necessitates the precise modeling of muscle and fur dynamics, a process that remains both labor-intensive and computationally expensive within traditional production workflows. While generative diffusion models have shown promise in diverse artistic workflows, their capacity for high-fidelity animal simulation remains largely unexploited. We present MoZoo, a generative dynamics solver that bypasses conventional refinement to synthesize high-fidelity animal videos from coarse meshes under multimodal guidance. We propose Role-Aware RoPE (RAR-RoPE) which employs role-based index remapping to synchronize motion alignment while decoupling reference information via fixed temporal offsets. Complementing this, Asymmetric Decoupled Attention partitions the latent sequence to enforce a unidirectional information flow, effectively preventing feature interference and improving computational efficiency. To address the scarcity of high-quality training data, we introduce MoZoo-Data, a synthetic-to-real pipeline that leverages a rendering engine and an inverse mapping approach to construct a large-scale dataset of paired sequences. Furthermore, we establish MoZooBench, a comprehensive benchmark with 120 mesh-video pairs. Experimental results demonstrate that MoZoo achieves high-fidelity fur simulation across diverse animal skeletons and layouts, preserving superior temporal and structural consistency.

---

### Playing Games with My Heart: An Evaluation of AI Companion Apps

Authors: Maribeth Rauh, Dick A. H. Blankvoort, Matias Duran, Caoilfhionn Ní Dheoráin, Harshvardhan J. Pandit, Syrine Enneifer, Siddharth D. Jaiswal, Anthony Ventresque, Abeba Birhane
Published: 2026-04-08
Categories: cs.CY, cs.AI, cs.HC
arXiv: https://arxiv.org/abs/2605.08093v2

The use of chatbots for various forms of companionship is growing rapidly, raising a myriad of questions about simulated relationships, emotional dependence, and psychological harm. While major platforms such as ChatGPT, Grok, and Character AI are the subject of a growing body of research and legal inquiries, apps explicitly built for simulating intimate interpersonal relationships remain under-explored. In this work, we evaluate the five most popular AI companion mobile applications for factors that encourage parasocial interaction and may manipulate users. We do this by manually annotating the user experience each offers. Specifically, we systematically record and quantify design dark patterns, anthropomorphism, stereotypes, erotica, and technical performance issues. We find that all apps contain substantial dark patterns aimed at increasing monetisation and user engagement. Erotica and gamification such as levelling are also prevalent, and although features vary considerably between applications, all apps have highly anthropomorphic design. These findings shed light on the mechanics used to leverage users' simulated relationships. On that basis, we put forward recommendations for regulators to strengthen consumer protection in this rapidly emerging market. Content warning: This article contains objectifying images of women, erotic images, textual references to incest, and other potentially sensitive, offensive, and distressing text.

---

### Playing Games with My Heart: An Evaluation of AI Companion Apps

Authors: Maribeth Rauh, Dick A. H. Blankvoort, Matias Duran, Caoilfhionn Ní Dheoráin, Harshvardhan J. Pandit, Siddharth D. Jaiswal, Anthony Ventresque, Abeba Birhane
Published: 2026-04-08
Categories: cs.CY, cs.AI, cs.HC
arXiv: https://arxiv.org/abs/2605.08093v1

The use of chatbots for various forms of companionship is growing rapidly, raising a myriad of questions about simulated relationships, emotional dependence, and psychological harm. While major platforms such as ChatGPT, Grok, and Character.AI are the subject of a growing body of research and legal inquiries, apps explicitly built for simulating intimate interpersonal relationships remain under-explored. In this work, we evaluate the five most popular AI companion mobile applications in the EU and UK markets for factors that encourage parasocial interaction and may manipulate users. We do this by manually annotating the user experience each offers. Specifically, we systematically record and quantify design dark patterns, anthropomorphism, stereotypes, erotica, and technical performance issues. We find that all apps contain substantial dark patterns aimed at increasing monetisation and user engagement. Erotica and gamification features such as levelling are also prevalent, and although other features vary considerably between applications, all apps have highly anthropomorphic design. These findings shed light on the mechanics used to leverage users' simulated relationships. On that basis, we put forward concrete recommendations for regulators to strengthen consumer protection in this rapidly emerging market. Content warning: This article contains objectifying images of women, erotic images, textual references to incest, and other potentially sensitive, offensive, and distressing text.

---