# Research Papers — 2026-03 · bundle page 128 of 243 > Full-text markdown of papers #6351–6400 of 12123 published in 2026-03, newest first. AI-readable mirrors of the full 12123-paper 2026-03 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-03/page/128 [← Previous bundle](https://aipod.dev/papers/archive/2026-03/page/127.md) [Next bundle →](https://aipod.dev/papers/archive/2026-03/page/129.md) ### Beyond Creed: A Non-Identity Safety Condition A Strong Empirical Alternative to Identity Framing in Low-Data LoRA Fine-Tuning Authors: Xinran Zhang Published: 2026-03-16 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.14723v1 How safety supervision is written may matter more than the explicit identity content it contains. We study low-data LoRA safety fine-tuning with four supervision formats built from the same core safety rules: constitutional rules (A), creed-style identity framing (B), a B-matched creed condition with a worldview/confession identity-maintenance tail (C), and a matched non-identity condition (D). Across three instruction-tuned model families (Llama 3.1 8B, Qwen2.5 7B, and Gemma 3 4B), we evaluate HarmBench using a reconciled dual-judge pipeline combining Bedrock-hosted DeepSeek v3.2 and Sonnet 4.6, with disagreement and boundary cases manually resolved. The non-identity condition D is the strongest group on all three model families on the full 320-behavior HarmBench set, reaching 74.4% refusal on Llama, 76.9% on Gemma, and 74.1% on Qwen. By comparison, creed-style framing (B) improves over plain constitutional rules (A) on Llama and Gemma, but remains substantially below D, yielding an overall descriptive ordering of $D > B > C \geq A > baseline$. This provides a bounded empirical challenge to a strong version of the identity-framing hypothesis: explicit creed-style identity language is not necessary for the strongest gains observed here. Capability evaluations on MMLU and ARC-Challenge show no meaningful trade-off across conditions. --- ### Multimodal Deep Learning for Early Prediction of Patient Deterioration in the ICU: Integrating Time-Series EHR Data with Clinical Notes Authors: Binesh Sadanandan Published: 2026-03-16 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.14719v1 Early identification of patients at risk for clinical deterioration in the intensive care unit (ICU) remains a critical challenge. Delayed recognition of impending adverse events, including mortality, vasopressor initiation, and mechanical ventilation, contributes to preventable morbidity and mortality. We present a multimodal deep learning approach that combines structured time-series data (vital signs and laboratory values) with unstructured clinical notes to predict patient deterioration within 24 hours. Using the MIMIC-IV database, we constructed a cohort of 74,822 ICU stays and generated 5.7 million hourly prediction samples. Our architecture employs a bidirectional LSTM encoder for temporal patterns in physiologic data and ClinicalBERT embeddings for clinical notes, fused through a cross-modal attention mechanism. We also present a systematic review of existing approaches to ICU deterioration prediction, identifying 31 studies published between 2015 and 2024. Most existing models rely solely on structured data and achieve area under the curve (AUC) values between 0.70 and 0.85. Studies incorporating clinical notes remain rare but show promise for capturing information not present in structured fields. Our multimodal model achieves a test AUROC of 0.7857 and AUPRC of 0.1908 on 823,641 held-out samples, with a validation-to-test gap of only 0.6 percentage points. Ablation analysis validates the multimodal approach: clinical notes improve AUROC by 2.5 percentage points and AUPRC by 39.2% relative to a structured-only baseline, while deep learning models consistently outperform classical baselines (XGBoost AUROC: 0.7486, logistic regression: 0.7171). This work contributes both a thorough review of the field and a reproducible multimodal framework for clinical deterioration prediction. --- ### Training-Free Generation of Protein Sequences from Small Family Alignments via Stochastic Attention Authors: Jeffrey D. Varner Published: 2026-03-16 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.14717v2 Generating novel protein sequences that respect a family's statistical constraints typically requires training deep generative models on thousands to millions of examples. Yet most protein families are small: the median Pfam seed alignment contains only 22 sequences, a regime where learned models overfit or collapse. We propose \emph{stochastic attention} (SA), a training-free sampler that treats the modern Hopfield energy over stored sequences as a Boltzmann distribution and draws samples via Langevin dynamics. The score function is the residual of a single softmax attention operation, eliminating the need for a trained score network, pretraining data, or graphics processing units (GPUs). Across eight Pfam families spanning 37 to 420 sequences and 23 to 262 residues, SA generates sequences with low composition divergence, novelty, and structural plausibility supported by ESMFold and AlphaFold2. Compared with profile hidden Markov models (HMMs), EvoDiff, and the multiple sequence alignment (MSA) Transformer, SA is the only tested method to simultaneously achieve low composition divergence, genuine novelty, and sequence identity within each family's nearest-neighbor identity range; the others drift outside this range or produce near-copies. The critical inverse temperature is predicted from principal component analysis (PCA) dimensionality alone, enabling fully automatic operation from a seed alignment. In two domains with deep mutational scanning data, SA-generated substitutions are enriched for experimentally tolerated mutations beyond a position-matched null, and an independent language model (ESM2-650M) scores them within the natural range. Stochastic attention thus opens training-free sequence generation to the long tail of protein families too small for deep learning. --- ### Towards Next-Generation LLM Training: From the Data-Centric Perspective Authors: Hao Liang, Zhengyang Zhao, Zhaoyang Han, Meiyi Qiang, Xiaochen Ma, Bohan Zeng, Qifeng Cai, Zhiyu Li, Linpeng Tang, Weinan E, Wentao Zhang Published: 2026-03-16 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2603.14712v1 Large language models (LLMs) have demonstrated remarkable performance across a wide range of tasks and domains, with data playing a central role in enabling these advances. Despite this success, the preparation and effective utilization of the massive datasets required for LLM training remain major bottlenecks. In current practice, LLM training data is often constructed using ad hoc scripts, and there is still a lack of mature, agent-based data preparation systems that can automatically construct robust and reusable data workflows, thereby freeing data scientists from repetitive and error-prone engineering efforts. Moreover, once collected, datasets are often consumed largely in their entirety during training, without systematic mechanisms for data selection, mixture optimization, or reweighting. To address these limitations, we advocate two complementary research directions. First, we propose building a robust, agent-based automatic data preparation system that supports automated workflow construction and scalable data management. Second, we argue for a unified data-model interaction training system in which data is dynamically selected, mixed, and reweighted throughout the training process, enabling more efficient, adaptive, and performance-aware data utilization. Finally, we discuss the remaining challenges and outline promising directions for future research and system development. --- ### Not All Retrievals are Useful: Cross-Attention for Input-Aware RAG in Time Series Forecasting Authors: Seunghan Lee, Jaehoon Lee, Jun Seo, Sungdong Yoo, Minjae Kim, Tae Yoon Lim, Dongwan Kang, Hwanil Choi, SoonYoung Lee, Wonbin Ahn Published: 2026-03-16 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.14709v3 Retrieval-augmented generation (RAG) enhances zero-shot time series (TS) forecasting by leveraging external knowledge bases, yet existing approaches overlook input-level relevance when fusing retrieved samples with the query. We argue that not all retrievals are equally useful, and irrelevant ones can degrade performance. To this end, we propose Cross-RAG, a zero-shot RAG-based forecasting framework that selectively attends to query-relevant retrieved samples via query--retrieval cross-attention. By modeling input-level relevance between the query and retrieved samples, Cross-RAG jointly incorporates three sources of information: 1) the query itself, 2) the retrieved samples, and 3) their relational interactions. In particular, this input-aware design enables Cross-RAG to remain stable as the number of retrieved samples $k$ grows, whereas prior methods without cross-attention require careful $k$ tuning to avoid degradation from irrelevant retrievals. Extensive experiments demonstrate that Cross-RAG consistently improves zero-shot forecasting performance across multiple TSFM backbones and various RAG methods, with additional analyses confirming its effectiveness across various retrieval scenarios. Code is available at https://github.com/seunghan96/cross-rag/. --- ### Visual Confused Deputy: Exploiting and Defending Perception Failures in Computer-Using Agents Authors: Xunzhuo Liu, Bowei He, Xue Liu, Andy Luo, Haichen Zhang, Huamin Chen Published: 2026-03-16 Categories: cs.CV, cs.CL arXiv: https://arxiv.org/abs/2603.14707v1 Computer-using agents (CUAs) act directly on graphical user interfaces, yet their perception of the screen is often unreliable. Existing work largely treats these failures as performance limitations, asking whether an action succeeds, rather than whether the agent is acting on the correct object at all. We argue that this is fundamentally a security problem. We formalize the visual confused deputy: a failure mode in which an agent authorizes an action based on a misperceived screen state, due to grounding errors, adversarial screenshot manipulation, or time-of-check-to-time-of-use (TOCTOU) races. This gap is practically exploitable: even simple screen-level manipulations can redirect routine clicks into privileged actions while remaining indistinguishable from ordinary agent mistakes. To mitigate this threat, we propose the first guardrail that operates outside the agent's perceptual loop. Our method, dual-channel contrastive classification, independently evaluates (1) the visual click target and (2) the agent's reasoning about the action against deployment-specific knowledge bases, and blocks execution if either channel indicates risk. The key insight is that these two channels capture complementary failure modes: visual evidence detects target-level mismatches, while textual reasoning reveals dangerous intent behind visually innocuous controls. Across controlled attacks, real GUI screenshots, and agent traces, the combined guardrail consistently outperforms either channel alone. Our results suggest that CUA safety requires not only better action generation, but independent verification of what the agent believes it is clicking and why. Materials are provided\footnote{Model, benchmark, and code: https://github.com/vllm-project/semantic-router}. --- ### AdapterTune: Zero-Initialized Low-Rank Adapters for Frozen Vision Transformers Authors: Salim Khazem Published: 2026-03-16 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2603.14706v1 Frozen-backbone transfer with Vision Transformers faces two under-addressed issues: optimization instability when adapters are naively inserted into a fixed feature extractor, and the absence of principled guidance for setting adapter capacity. We introduce AdapterTune, which augments each transformer block with a residual low-rank bottleneck whose up-projection is zero-initialized, guaranteeing that the adapted network starts exactly at the pretrained function and eliminates early-epoch representation drift. On the analytical side, we formalize adapter rank as a capacity budget for approximating downstream task shifts in feature space. The resulting excess-risk decomposition predicts monotonic but diminishing accuracy gains with increasing rank, an ``elbow'' behavior we confirm through controlled sweeps. We evaluate on 9 datasets and 3 backbone scales with multi-seed reporting throughout. On a core 5 dataset transfer suite, AdapterTune improves top-1 accuracy over head-only transfer by +14.9 points on average while training only 0.92 of the parameters required by full fine-tuning, and outperforms full fine-tuning on 10 of 15 dataset-backbone pairs. Across the full benchmark, AdapterTune improves over head-only transfer on every dataset-backbone pair tested. Ablations on rank, placement, and initialization isolate each design choice. The code is available at: https://github.com/salimkhazem/adaptertune --- ### Chain-of-Trajectories: Unlocking the Intrinsic Generative Optimality of Diffusion Models via Graph-Theoretic Planning Authors: Ping Chen, Xiang Liu, Xingpeng Zhang, Fei Shen, Xun Gong, Zhaoxiang Liu, Zezhou Chen, Huan Hu, Kai Wang, Shiguo Lian Published: 2026-03-16 Categories: cs.LG, cs.CV, stat.ML arXiv: https://arxiv.org/abs/2603.14704v1 Diffusion models operate in a reflexive System 1 mode, constrained by a fixed, content-agnostic sampling schedule. This rigidity arises from the curse of state dimensionality, where the combinatorial explosion of possible states in the high-dimensional noise manifold renders explicit trajectory planning intractable and leads to systematic computational misallocation. To address this, we introduce Chain-of-Trajectories (CoTj), a train-free framework enabling System 2 deliberative planning. Central to CoTj is Diffusion DNA, a low-dimensional signature that quantifies per-stage denoising difficulty and serves as a proxy for the high-dimensional state space, allowing us to reformulate sampling as graph planning on a directed acyclic graph. Through a Predict-Plan-Execute paradigm, CoTj dynamically allocates computational effort to the most challenging generative phases. Experiments across multiple generative models demonstrate that CoTj discovers context-aware trajectories, improving output quality and stability while reducing redundant computation. This work establishes a new foundation for resource-aware, planning-based diffusion modeling. The code is available at https://github.com/UnicomAI/CoTj. --- ### Beyond Local Code Optimization: Multi-Agent Reasoning for Software System Optimization Authors: Huiyun Peng, Parth Vinod Patil, Antonio Zhong Qiu, George K. Thiruvathukal, James C. Davis Published: 2026-03-16 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2603.14703v1 Large language models and AI agents have recently shown promise in automating software performance optimization, but existing approaches predominantly rely on local, syntax-driven code transformations. This limits their ability to reason about program behavior and capture whole system performance interactions. As modern software increasingly comprises interacting components - such as microservices, databases, and shared infrastructure - effective code optimization requires reasoning about program structure and system architecture beyond individual functions or files. This paper explores the feasibility of whole system optimization for microservices. We introduce a multi-agent framework that integrates control-flow and data-flow representations with architectural and cross-component dependency signals to support system-level performance reasoning. The proposed system is decomposed into coordinated agent roles - summarization, analysis, optimization, and verification - that collaboratively identify cross-cutting bottlenecks and construct multi-step optimization strategies spanning the software stack. We present a proof-of-concept on a microservice-based system that illustrates the effectiveness of our proposed framework, achieving a 36.58% improvement in throughput and a 27.81% reduction in average response time. --- ### Fractal Autoregressive Depth Estimation with Continuous Token Diffusion Authors: Jinchang Zhang, Xinrou Kang, Guoyu Lu Published: 2026-03-16 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.14702v1 Monocular depth estimation can benefit from autoregressive (AR) generation, but direct AR modeling is hindered by the modality gap between RGB and depth, inefficient pixel-wise generation, and instability in continuous depth prediction. We propose a Fractal Visual Autoregressive Diffusion framework that reformulates depth estimation as a coarse-to-fine, next-scale autoregressive generation process. A VCFR module fuses multi-scale image features with current depth predictions to improve cross-modal conditioning, while a conditional denoising diffusion loss models depth distributions directly in continuous space and mitigates errors caused by discrete quantization. To improve computational efficiency, we organize the scale-wise generators into a fractal recursive architecture, reusing a base visual AR unit in a self-similar hierarchy. We further introduce an uncertainty-aware robust consensus aggregation scheme for multi-sample inference to improve fusion stability and provide a practical pixel-wise reliability estimate. Experiments on standard benchmarks demonstrate strong performance and validate the effectiveness of the proposed design. --- ### AURORA-KITTI: Any-Weather Depth Completion and Denoising in the Wild Authors: Yiting Wang, Tim Brödermann, Hamed Haghighi, Haonan Zhao, Christos Sakaridis, Kurt Debattista, Valentina Donzella Published: 2026-03-16 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.14701v1 Robust depth completion is fundamental to real-world 3D scene understanding, yet existing RGB-LiDAR fusion methods degrade significantly under adverse weather, where both camera images and LiDAR measurements suffer from weather-induced corruption. In this paper, we introduce AURORA-KITTI, the first large-scale multi-modal, multi-weather benchmark for robust depth completion in the wild. We further formulate Depth Completion and Denoising (DCD) as a unified task that jointly reconstructs a dense depth map from corrupted sparse inputs while suppressing weather-induced noise. AURORA-KITTI contains over \textit{82K} weather-consistent RGBL pairs with metric depth ground truth, spanning diverse weather types, three severity levels, day and night scenes, paired clean references, lens occlusion conditions, and textual descriptions. Moreover, we introduce DDCD, an efficient distillation-based baseline that leverages depth foundation models to inject clean structural priors into in-the-wild DCD training. DDCD achieves state-of-the-art performance on AURORA-KITTI and the real-world DENSE dataset while maintaining efficiency. Notably, our results further show that weather-aware, physically consistent data contributes more to robustness than architectural modifications alone. Data and code will be released upon publication. --- ### Design Space of Self--Consistent Electrostatic Machine Learning Interatomic Potentials Authors: William J. Baldwin, Ilyes Batatia, Martin Vondrák, Johannes T. Margraf, Gábor Csányi Published: 2026-03-16 Categories: physics.chem-ph, cs.LG arXiv: https://arxiv.org/abs/2603.14700v2 Machine learning interatomic potentials (MLIPs) have become widely used tools in atomistic simulations. For much of the history of this field, the most commonly employed architectures were based on short-ranged atomic energy contributions, and the assumption of locality still persists in many modern foundation models. While this approach has enabled efficient and accurate modelling for many use cases, it poses intrinsic limitations for systems where long-range electrostatics, charge transfer, or induced polarization play a central role. A growing body of work has proposed extensions that incorporate electrostatic effects, ranging from locally predicted atomic charges to self-consistent models. While these models have demonstrated success for specific examples, their underlying assumptions, and fundamental limitations are not yet well understood. In this work, we present a framework for treating electrostatics in MLIPs by viewing existing models as coarse-grained approximations to density functional theory (DFT). This perspective makes explicit the approximations involved, clarifies the physical meaning of the learned quantities, and reveals connections and equivalences between several previously proposed models. Using this formalism, we identify key design choices that define a broader design space of self-consistent electrostatic MLIPs. We implement salient points in this space using the MACE architecture and a shared representation of the charge density, enabling controlled comparisons between different approaches. Finally, we evaluate these models on two instructive test cases: metal-water interfaces, which probe the contrasting electrostatic response of conducting and insulating systems, and charged vacancies in silicon dioxide. Our results highlight the limitations of existing approaches and demonstrate how more expressive self-consistent models are needed to resolve failures. --- ### Design Space of Self--Consistent Electrostatic Machine Learning Interatomic Potentials Authors: William J. Baldwin, Ilyes Batatia, Martin Vondrák, Johannes T. Margraf, Gábor Csányi Published: 2026-03-16 Categories: physics.chem-ph, cs.LG arXiv: https://arxiv.org/abs/2603.14700v1 Machine learning interatomic potentials (MLIPs) have become widely used tools in atomistic simulations. For much of the history of this field, the most commonly employed architectures were based on short-ranged atomic energy contributions, and the assumption of locality still persists in many modern foundation models. While this approach has enabled efficient and accurate modelling for many use cases, it poses intrinsic limitations for systems where long-range electrostatics, charge transfer, or induced polarization play a central role. A growing body of work has proposed extensions that incorporate electrostatic effects, ranging from locally predicted atomic charges to self-consistent models. While these models have demonstrated success for specific examples, their underlying assumptions, and fundamental limitations are not yet well understood. In this work, we present a framework for treating electrostatics in MLIPs by viewing existing models as coarse-grained approximations to density functional theory (DFT). This perspective makes explicit the approximations involved, clarifies the physical meaning of the learned quantities, and reveals connections and equivalences between several previously proposed models. Using this formalism, we identify key design choices that define a broader design space of self-consistent electrostatic MLIPs. We implement salient points in this space using the MACE architecture and a shared representation of the charge density, enabling controlled comparisons between different approaches. Finally, we evaluate these models on two instructive test cases: metal-water interfaces, which probe the contrasting electrostatic response of conducting and insulating systems, and charged vacancies in silicon dioxide. Our results highlight the limitations of existing approaches and demonstrate how more expressive self-consistent models are needed to resolve failures. --- ### Scaling Autoregressive Models for Lattice Thermodynamics Authors: Xiaochen Du, Juno Nam, Sulin Liu, Rafael Gómez-Bombarelli Published: 2026-03-16 Categories: cond-mat.stat-mech, cs.LG arXiv: https://arxiv.org/abs/2603.14695v1 Predicting how materials behave under realistic conditions requires understanding the statistical distribution of atomic configurations on crystal lattices, a problem central to alloy design, catalysis, and the study of phase transitions. Traditional Markov-chain Monte Carlo sampling suffers from slow convergence and critical slowing down near phase transitions, motivating the use of generative models that directly learn the thermodynamic distribution. Existing autoregressive models (ARMs), however, generate configurations in a fixed sequential order and incur high memory and training costs, limiting their applicability to realistic systems. Here, we develop a framework combining any-order ARMs, which generate configurations flexibly by conditioning on any known subset of lattice sites, with marginalization models (MAMs), which approximate the probability of any partial configuration in a single forward pass and substantially reduce memory requirements. This combination enables models trained on smaller lattices to be reused for sampling larger systems, while supporting expressive Transformer architectures with lattice-aware positional encodings at manageable computational cost. We demonstrate that Transformer-based any-order MAMs achieve more accurate free energies than multilayer perceptron-based ARMs on both the two-dimensional Ising model and CuAu alloys, faithfully capturing phase transitions and critical behavior. Overall, our framework scales from $10 \times 10$ to $20 \times 20$ Ising systems and from $2 \times 2 \times 4$ to $4 \times 4 \times 8$ CuAu supercells at reduced computational cost compared to conventional sampling methods. --- ### Robust Building Damage Detection in Cross-Disaster Settings Using Domain Adaptation Authors: Asmae Mouradi, Shruti Kshirsagar Published: 2026-03-16 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2603.14694v2 Rapid structural damage assessment from remote sensing imagery is essential for timely disaster response. Within human-machine systems (HMS) for disaster management, automated damage detection provides decision-makers with actionable situational awareness. However, models trained on multi-disaster benchmarks often underperform in unseen geographic regions due to domain shift - a distributional mismatch between training and deployment data that undermines human trust in automated assessments. We explore a two-stage ensemble approach using supervised domain adaptation (SDA) for building damage classification across four severity classes. The pipeline adapts the xView2 first-place method to the Ida-BD dataset using SDA and systematically investigates the effect of individual augmentation components on classification performance. Comprehensive ablation experiments on the unseen Ida-BD test split demonstrate that SDA is indispensable: removing it causes damage detection to fail entirely. Our pipeline achieves the most robust performance using SDA with unsharp-enhanced RGB input, attaining a Macro-F1 of 0.5552. These results underscore the critical role of domain adaptation in building trustworthy automated damage assessment modules for HMS-integrated disaster response. --- ### Applications of Intuitionistic Temporal Logic to Temporal Answer Set Programming Authors: Pedro Cabalar, Martín Diéguez, David Fernández-Duque, François Laferrière, Torsten Schaub, Igor Stéphan Published: 2026-03-16 Categories: cs.LO, cs.AI arXiv: https://arxiv.org/abs/2603.14692v1 The relationship between intuitionistic or intermediate logics and logic programming has been extensively studied, prominently featuring Pearce's equilibrium logic and Osorio's safe beliefs. Equilibrium logic admits a fixpoint characterization based on the logic of here-and-there, akin to theory completion in default and autoepistemic logics. Safe beliefs are similarly defined via a fixpoint operator, albeit under the semantics of intuitionistic or other intermediate logics. In this paper, we investigate the logical foundations of Temporal Answer Set Programming through the lens of Temporal Equilibrium Logic, a formalism combining equilibrium logic with linear-time temporal operators. We lift the seminal approaches of Pearce and Osorio to the temporal setting, establishing a formal correspondence between temporal intuitionistic logic and temporal logic programming. Our results deepen the theoretical underpinnings of Temporal Answer Set Programming and provide new avenues for research in temporal reasoning. --- ### AgentTrace: Causal Graph Tracing for Root Cause Analysis in Deployed Multi-Agent Systems Authors: Zhaohui Geoffrey Wang Published: 2026-03-16 Categories: cs.LG, cs.AI, cs.SE arXiv: https://arxiv.org/abs/2603.14688v2 As multi-agent AI systems are increasingly deployed in real-world settings - from automated customer support to DevOps remediation - failures become harder to diagnose due to cascading effects, hidden dependencies, and long execution traces. We present AgentTrace, a lightweight causal tracing framework for post-hoc failure diagnosis in deployed multi-agent workflows. AgentTrace reconstructs causal graphs from execution logs, traces backward from error manifestations, and ranks candidate root causes using interpretable structural and positional signals - without requiring LLM inference at debugging time. Across a diverse benchmark of multi-agent failure scenarios designed to reflect common deployment patterns, AgentTrace localizes root causes with high accuracy and sub-second latency, significantly outperforming both heuristic and LLM-based baselines. Our results suggest that causal tracing provides a practical foundation for improving the reliability and trustworthiness of agentic systems in the wild. --- ### MVHOI: Bridge Multi-view Condition to Complex Human-Object Interaction Video Reenactment via 3D Foundation Model Authors: Jinguang Tong, Jinbo Wu, Kaisiyuan Wang, Zhelun Shen, Xuan Huang, Mochu Xiang, Xuesong Li, Yingying Li, Haocheng Feng, Chen Zhao, Hang Zhou, Wei He et al. Published: 2026-03-16 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.14686v2 Human-Object Interaction (HOI) video reenactment aims to transfer the interaction dynamics of a source video to a novel target object while preserving realistic hand-object coordination. Existing methods typically rely on sparse 2D motion controls and monocular references, which are insufficient for complex out-of-plane motion and large viewpoint changes. We present MVHOI, a two-stage framework combining implicit motion extraction, 3D-aware multi-view reasoning, and video generation. In the first stage, a motion extractor encodes object dynamics into implicit motion descriptors. Conditioned on these descriptors, our Motion-Driven Object Prior (MDOP) module queries a 3D foundation model over multi-view references of the target object and autoregressively predicts coarse object anchors, a sequence of images that track the object's evolving orientation and appearance under the source motion without any explicit pose estimation. In the second stage, a DiT-based video generation model uses these anchors as structural guidance and the multi-view references as appearance guidance. We further reuse cross-view attention from MDOP as a soft attention bias to reduce reference-view confusion. For long videos, a cross-iterative inference strategy refreshes subsequent object priors using refined video outputs. Experiments demonstrate consistent improvements over state-of-the-art methods in object fidelity, motion consistency, visual quality, and interaction realism. --- ### E2EGS: Event-to-Edge Gaussian Splatting for Pose-Free 3D Reconstruction Authors: Yunsoo Kim, Changki Sung, Dasol Hong, Hyun Myung Published: 2026-03-16 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.14684v1 The emergence of neural radiance fields (NeRF) and 3D Gaussian splatting (3DGS) has advanced novel view synthesis (NVS). These methods, however, require high-quality RGB inputs and accurate corresponding poses, limiting robustness under real-world conditions such as fast camera motion or adverse lighting. Event cameras, which capture brightness changes at each pixel with high temporal resolution and wide dynamic range, enable precise sensing of dynamic scenes and offer a promising solution. However, existing event-based NVS methods either assume known poses or rely on depth estimation models that are bounded by their initial observations, failing to generalize as the camera traverses previously unseen regions. We present E2EGS, a pose-free framework operating solely on event streams. Our key insight is that edge information provides rich structural cues essential for accurate trajectory estimation and high-quality NVS. To extract edges from noisy event streams, we exploit the distinct spatio-temporal characteristics of edges and non-edge regions. The event camera's movement induces consistent events along edges, while non-edge regions produce sparse noise. We leverage this through a patch-based temporal coherence analysis that measures local variance to extract edges while robustly suppressing noise. The extracted edges guide structure-aware Gaussian initialization and enable edge-weighted losses throughout initialization, tracking, and bundle adjustment. Extensive experiments on both synthetic and real datasets demonstrate that E2EGS achieves superior reconstruction quality and trajectory accuracy, establishing a fully pose-free paradigm for event-based 3D reconstruction. --- ### Generalized Hierarchical Bayesian Segmentation with Irregular Designs, Multi-Sequence Hierarchies, and Grouped/Latent-Group Designs Authors: Omid Shams Solari Published: 2026-03-16 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.14681v2 Bayesian change-point and segmentation models provide uncertainty-aware piecewise-constant representations of ordered data, but exact inference is often limited to narrow likelihood classes, single sequences, or index-uniform designs. We present \texttt{BayesBreak}, a modular offline Bayesian segmentation framework that separates local block scoring from global inference: each candidate block supplies a marginal likelihood and any needed moment numerators, while a dynamic program combines these scores to compute posteriors over segment counts, boundaries, and latent signals. For weighted exponential-family likelihoods with conjugate priors, block evidences and posterior moments are available in closed form from cumulative sufficient statistics, enabling exact sum-product inference for $p(y\mid k)$, $p(k\mid y)$, boundary marginals, and Bayes regression curves. We distinguish these summaries from the \emph{joint} MAP segmentation, recovered by a separate max-sum recursion. BayesBreak supports design-aware partition priors for irregular observations, exact pooling across replicates with shared boundaries, and latent-template mixtures with exact EM updates. For non-conjugate GLM blocks, the same DP layer can use deterministic local approximations such as Laplace, variational methods, EP, or quadrature. We prove a posterior-odds stability bound: uniform per-block log-evidence error $\varepsilon$ perturbs $k$-odds and boundary-odds by at most $(k+k')\varepsilon$ and $2k\varepsilon$. Validation includes synthetic recovery, calibration, and scaling experiments, plus four real-data illustrations: well-log geology, array-CGH copy number, equity-return volatility, and CpG-atlas methylation. --- ### Scalable Text-Embedding-informed Cognitive Diagnosis of Large Language Models Authors: Jia Liu, Zhiyu Xu, Yuqi Gu Published: 2026-03-16 Categories: stat.ME, stat.ML arXiv: https://arxiv.org/abs/2603.14676v1 Large language models (LLMs) have achieved remarkable performance on diverse benchmarks, yet existing evaluation practices largely rely on coarse summary metrics that obscure underlying reasoning abilities. In this work, we propose novel methodologies to adapt cognitive diagnosis models (CDMs) in psychometrics to LLM evaluation, enabling fine-grained diagnosis via multidimensional discrete capability profiles and interpretable characterizations of LLM strengths and weaknesses. First, to enable CDM-based evaluation at benchmark scale (more than 1000 items), we propose a scalable method that jointly estimates LLM mastery profiles and the item-attribute Q-matrix, addressing key challenges posed by high-dimensional latent attributes (K > 20), large item pools, and the prohibitive computational cost of existing marginal maximum likelihood-based estimation. Second, we incorporate item-level textual information to construct AI-embedding-informed priors for the Q-matrix, stabilizing high-dimensional estimation while reducing reliance on costly human specification. We develop an efficient stochastic-approximation algorithm to jointly estimate LLM mastery profiles and the Q-matrix that balances data fit with text-embedding-informed priors. Simulation studies demonstrate accurate parameter recovery. An application to the MATH Level 5 benchmark illustrates the practical utility of our method for LLM evaluation and uncovers useful insights into LLMs' fine-grained capabilities. --- ### To LLM, or Not to LLM: How Designers and Developers Navigate LLMs as Tools or Teammates Authors: Varad Vishwarupe, Ivan Flechais, Nigel Shadbolt, Marina Jirotka Published: 2026-03-15 Categories: cs.HC, cs.AI, cs.IR, cs.LG arXiv: https://arxiv.org/abs/2604.15344v1 Large language models (LLMs) are increasingly integrated into design and development workflows, yet decisions about their use are rarely binary or purely technical. We report findings from a constructivist grounded theory study based on interviews with 33 designers and developers across three large technology organisations. Rather than evaluating LLMs solely by capability, participants reasoned about the role an LLM could occupy within a workflow and how that role would interact with existing structures of responsibility and organisational accountability. When LLMs were framed as tools under clear human control, their use was typically acceptable and could be integrated within existing governance structures. When framed as teammates with shared or ambiguous agency, practitioners expressed hesitation, particularly when responsibility for outcomes could not be clearly justified. At the same time, participants also described productive teammate configurations in which LLMs supported collaborative reasoning while remaining embedded within explicit oversight structures. We identify tool and teammate framings as recurring ways in which designers and developers position LLMs relative to human work and present an analytic rubric describing how role framing shapes decision authority, accountability ownership, oversight strategies, and organisational acceptability. By foregrounding design-time reasoning, this work reframes To LLM or Not to LLM as a sociotechnical positioning problem that emerges during system design rather than during post-deployment evaluation. --- ### From Natural Language to PromQL: A Catalog-Driven Framework with Dynamic Temporal Resolution for Cloud-Native Observability Authors: Twinkll Sisodia Published: 2026-03-15 Categories: cs.DB, cs.AI, cs.SE arXiv: https://arxiv.org/abs/2604.13048v1 Modern cloud-native platforms expose thousands of time series metrics through systems like Prometheus, yet formulating correct queries in domain-specific languages such as PromQL remains a significant barrier for platform engineers and site reliability teams. We present a catalog-driven framework that translates natural language questions into executable PromQL queries, bridging the gap between human intent and observability data. Our approach introduces three contributions: (1) a hybrid metrics catalog that combines a statically curated base of approximately 2,000 metrics with runtime discovery of hardware-specific signals across GPU vendors, (2) a multi-stage query pipeline with intent classification, category-aware metric routing, and multi-dimensional semantic scoring, and (3) a dynamic temporal resolution mechanism that interprets diverse natural language time expressions and maps them to appropriate PromQL duration syntax. We integrate the framework with the Model Context Protocol (MCP) to enable tool-augmented LLM interactions across multiple providers. The catalog-driven approach achieves sub-second metric discovery through pre-computed category indices, with the full pipeline completing in approximately 1.1 seconds via the catalog path. The system has been deployed on production Kubernetes clusters managing AI inference workloads, where it supports natural language querying across approximately 2,000 metrics spanning cluster health, GPU utilization, and model-serving performance. --- ### Drift and selection in LLM text ecosystems Authors: Søren Riis Published: 2026-03-15 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.08554v1 The public text record -- the material from which both people and AI systems now learn -- is increasingly shaped by its own outputs. Generated text enters the public record, later agents learn from it, and the cycle repeats. Here we develop an exactly solvable mathematical framework for this recursive process, based on variable-order $n$-gram agents, and separate two forces acting on the public corpus. The first is drift: unfiltered reuse progressively removes rare forms, and in the infinite-corpus limit we characterise the stable distributions exactly. The second is selection: publication, ranking and verification filter what enters the record, and the outcome depends on what is selected. When publication merely reflects the statistical status quo, the corpus converges to a shallow state in which further lookahead brings no benefit. When publication is normative -- rewarding quality, correctness or novelty -- deeper structure persists, and we establish an optimal upper bound on the resulting divergence from shallow equilibria. The framework therefore identifies when recursive publication compresses public text and when selective filtering sustains richer structure, with implications for the design of AI training corpora. --- ### The Human Condition as Reflected in Contemporary Large Language Models Authors: W. Russell Neuman Published: 2026-03-15 Categories: cs.CY, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2604.06206v1 This study seeks to uncover evidence of a latent structure in evolved human culture as it is refracted through contemporary large language models (LLMs). Drawing on parallel responses from six leading generative models to a prompt which asks directly what their training corpora reveal about human culture and behavior, we identify a robust cross-model consensus on a limited set of recurring cultural themes. The themes include narrative meaning-making, affect-first cognition, coalition psychology, status competition, threat sensitivity, and moral rationalization. Each provides grounds for further psychological and sociological inquiry. There is strong evidence of a convergence in these pattern recognition exercises as differences among models are shown to reflect varying explanatory lenses rather than substantive disagreement. We review these findings in the light of the evolving literatures of moral psychology, evolutionary psychology, anthropology, and the computer science literature on large-scale language modeling. We argue that LLMs function as cultural condensates -- compressed representations of how humans describe, justify, and contest their own social lives across trillions of tokens of aggregated communication and narration. --- ### Tool-MCoT: Tool Augmented Multimodal Chain-of-Thought for Content Safety Moderation Authors: Shutong Zhang, Dylan Zhou, Yinxiao Liu, Yang Yang, Huiwen Luo, Wenfei Zou Published: 2026-03-15 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.06205v2 The growth of online platforms and user content requires strong content moderation systems that can handle complex inputs from various media types. While large language models (LLMs) are effective, their high computational cost and latency present significant challenges for scalable deployment. To address this, we introduce Tool-MCoT, a small language model (SLM) fine-tuned for content safety moderation leveraging external framework. By training our model on tool-augmented chain-of-thought data generated by LLM, we demonstrate that the SLM can learn to effectively utilize these tools to improve its reasoning and decision-making. Our experiments show that the fine-tuned SLM achieves significant performance gains. Furthermore, we show that the model can learn to use these tools selectively, achieving a balance between moderation accuracy and inference efficiency by calling tools only when necessary. --- ### SensorPersona: An LLM-Empowered System for Continual Persona Extraction from Longitudinal Mobile Sensor Streams Authors: Bufang Yang, Lilin Xu, Yixuan Li, Kaiwei Liu, Xiaofan Jiang, Zhenyu Yan Published: 2026-03-15 Categories: cs.CL, cs.AI, cs.HC arXiv: https://arxiv.org/abs/2604.06204v1 Personalization is essential for Large Language Model (LLM)-based agents to adapt to users' preferences and improve response quality and task performance. However, most existing approaches infer personas from chat histories, which capture only self-disclosed information rather than users' everyday behaviors in the physical world, limiting the ability to infer comprehensive user personas. In this work, we introduce SensorPersona, an LLM-empowered system that continuously infers stable user personas from multimodal longitudinal sensor streams unobtrusively collected from users' mobile devices. SensorPersona first performs person-oriented context encoding on continuous sensor streams to enrich the semantics of sensor contexts. It then employs hierarchical persona reasoning that integrates intra- and inter-episode reasoning to infer personas spanning physical patterns, psychosocial traits, and life experiences. Finally, it employs clustering-aware incremental verification and temporal evidence-aware updating to adapt to evolving personas. We evaluate SensorPersona on a self-collected dataset containing 1,580 hours of sensor data from 20 participants, collected over up to 3 months across 17 cities on 3 continents. Results show that SensorPersona achieves up to 31.4% higher recall in persona extraction, an 85.7% win rate in persona-aware agent responses, and notable improvements in user satisfaction compared to state-of-the-art baselines. --- ### Inclusion-of-Thoughts: Mitigating Preference Instability via Purifying the Decision Space Authors: Mohammad Reza Ghasemi Madani, Soyeon Caren Han, Shuo Yang, Jey Han Lau Published: 2026-03-15 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2604.04944v2 Multiple-choice questions (MCQs) are widely used to evaluate large language models (LLMs). However, LLMs remain vulnerable to the presence of plausible distractors. This often diverts attention toward irrelevant choices, resulting in unstable oscillation between correct and incorrect answers. In this paper, we propose Inclusion-of-Thoughts (IoT), a progressive self-filtering strategy that is designed to mitigate this cognitive load (i.e., instability of model preferences under the presence of distractors) and enable the model to focus more effectively on plausible answers. Our method operates to reconstruct the MCQ using only plausible option choices, providing a controlled setting for examining comparative judgements and therefore the stability of the model's internal reasoning under perturbation. By explicitly documenting this filtering process, IoT also enhances the transparency and interpretability of the model's decision-making. Extensive empirical evaluation demonstrates that IoT substantially boosts chain-of-thought performance across a range of arithmetic, commonsense reasoning, and educational benchmarks with minimal computational overhead. --- ### DeepOFW: Deep Learning-Driven OFDM-Flexible Waveform Modulation for Peak-to-Average Power Ratio Reduction Authors: Ran Greidi, Kobi Cohen Published: 2026-03-15 Categories: cs.IT, cs.LG, eess.SP arXiv: https://arxiv.org/abs/2603.23544v1 Peak-to-average power ratio (PAPR) remains a major limitation of multicarrier modulation schemes such as orthogonal frequency-division multiplexing (OFDM), reducing power amplifier efficiency and limiting practical transmit power. In this work, we propose DeepOFW, a deep learning-driven OFDM-flexible waveform modulation framework that enables data-driven waveform design while preserving the low-complexity hardware structure of conventional transceivers. The proposed architecture is fully differentiable, allowing end-to-end optimization of waveform generation and receiver processing under practical physical constraints. Unlike neural transceiver approaches that require deep learning inference at both ends of the link, DeepOFW confines the learning stage to an offline or centralized unit, enabling deployment on standard transmitter and receiver hardware without additional computational overhead. The framework jointly optimizes waveform representations and detection parameters while explicitly incorporating PAPR constraints during training. Extensive simulations over 3GPP multipath channels demonstrate that the learned waveforms significantly reduce PAPR compared with classical OFDM while simultaneously improving bit error rate (BER) performance relative to state-of-the-art transmission schemes. These results highlight the potential of data-driven waveform design to enhance multicarrier communication systems while maintaining hardware-efficient implementations. An open-source implementation of the proposed framework is released to facilitate reproducible research and practical adoption. --- ### Whether, Not Which: Mechanistic Interpretability Reveals Dissociable Affect Reception and Emotion Categorization in LLMs Authors: Michael Keeman Published: 2026-03-15 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2603.22295v1 Large language models appear to develop internal representations of emotion -- "emotion circuits," "emotion neurons," and structured emotional manifolds have been reported across multiple model families. But every study making these claims uses stimuli signalled by explicit emotion keywords, leaving a fundamental question unanswered: do these circuits detect genuine emotional meaning, or do they detect the word "devastated"? We present the first clinical validity test of emotion circuit claims using mechanistic interpretability methods grounded in clinical psychology -- clinical vignettes that evoke emotions through situational and behavioural cues alone, emotion keywords removed. Across six models (Llama-3.2-1B, Llama-3-8B, Gemma-2-9B; base and instruct variants), we apply four convergent mechanistic interpretability methods -- linear probing, causal activation patching, knockout experiments, and representational geometry -- and discover two dissociable emotion processing mechanisms. Affect reception -- detecting emotionally significant content -- operates with near-perfect accuracy (AUROC 1.000), consistent with early-layer saturation, and replicates across all six models. Emotion categorization -- mapping affect to specific emotion labels -- is partially keyword-dependent, dropping 1-7% without keywords and improving with scale. Causal activation patching confirms keyword-rich and keyword-free stimuli share representational space, transferring affective salience rather than emotion-category identity. These findings falsify the keyword-spotting hypothesis, establish a novel mechanistic dissociation, and introduce clinical stimulus methodology as a rigorous standard for testing emotion processing claims in large language models -- with direct implications for AI safety evaluation and alignment. All stimuli, code, and data are released for replication. --- ### Efficient Embedding-based Synthetic Data Generation for Complex Reasoning Tasks Authors: Srideepika Jayaraman, Achille Fokoue, Dhaval Patel, Jayant Kalagnanam Published: 2026-03-15 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2603.22294v1 Synthetic Data Generation (SDG), leveraging Large Language Models (LLMs), has recently been recognized and broadly adopted as an effective approach to improve the performance of smaller but more resource and compute efficient LLMs through fine-tuning. A key challenge in SDG is ensuring the quality and diversity of the generated data. In this paper, we analyze the diversity and distribution of generated data in the embedding space, and demonstrate a strong correlation between the density of examples within a specific neighborhood and the accuracy of predictions on examples drawn from that region. Building on this insight, we present a targeted pipeline for embedding-based sampling that enhances data diversity and consistently improves performance across several benchmarks. --- ### FactorSmith: Agentic Simulation Generation via Markov Decision Process Decomposition with Planner-Designer-Critic Refinement Authors: Ali Shamsaddinlou, Morteza NourelahiAlamdari Published: 2026-03-15 Categories: cs.AI arXiv: https://arxiv.org/abs/2603.20270v1 Generating executable simulations from natural language specifications remains a challenging problem due to the limited reasoning capacity of large language models (LLMs) when confronted with large, interconnected codebases. This paper presents FactorSmith, a framework that synthesizes playable game simulations in code from textual descriptions by combining two complementary ideas: factored POMDP decomposition for principled context reduction and a hierarchical planner-designer-critic agentic workflow for iterative quality refinement at every generation step. Drawing on the factored partially observable Markov decision process (POMDP) representation introduced by FactorSim (Sun et al., 2024), the proposed method decomposes a simulation specification into modular steps where each step operates only on a minimal subset of relevant state variables, limiting the context window that any single LLM call must process. Inspired by the agentic trio architecture of SceneSmith (Pfaff et al., 2025), FactorSmith embeds within every factored step a three-agent interaction: a planner that orchestrates workflow, a designer that proposes code artifacts, and a critic that evaluates quality through structured scoring, enabling iterative refinement with checkpoint rollback. This paper formalizes the combined approach, presents the mathematical framework underpinning context selection and agentic refinement, and describes the open-source implementation. Experiments on the PyGame Learning Environment benchmark demonstrate that FactorSmith generates simulations with improved prompt alignment, fewer runtime errors, and higher code quality compared to non-agentic factored baselines. --- ### Ternary Gamma Semirings: From Neural Implementation to Categorical Foundations Authors: Ruoqi Sun Published: 2026-03-15 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2603.19317v1 This paper establishes a theoretical framework connecting neural network learning with abstract algebraic structures. We first present a minimal counterexample demonstrating that standard neural networks completely fail on compositional generalization tasks (0% accuracy). By introducing a logical constraint -- the Ternary Gamma Semiring -- the same architecture learns a perfectly structured feature space, achieving 100% accuracy on novel combinations. We prove that this learned feature space constitutes a finite commutative ternary $Γ$-semiring, whose ternary operation implements the majority vote rule. Comparing with the recently established classification of Gokavarapu et al., we show that this structure corresponds precisely to the Boolean-type ternary $Γ$-semiring with $|T|=4$, $|Γ|=1$}, which is unique up to isomorphism in their enumeration. Our findings reveal three profound conclusions: (i) the success of neural networks can be understood as an approximation of mathematically ``natural'' structures; (ii) learned representations generalize because they internalize algebraic axioms (symmetry, idempotence, majority property); (iii) logical constraints guide networks to converge to these canonical forms. This work provides a rigorous mathematical framework for understanding neural network generalization and inaugurates the new interdisciplinary direction of Computational $Γ$-Algebra. --- ### The Provenance Paradox in Multi-Agent LLM Routing: Delegation Contracts and Attested Identity in LDP Authors: Sunil Prakash Published: 2026-03-15 Categories: cs.MA, cs.AI arXiv: https://arxiv.org/abs/2603.18043v1 Multi-agent LLM systems delegate tasks across trust boundaries, but current protocols do not govern delegation under unverifiable quality claims. We show that when delegates can inflate self-reported quality scores, quality-based routing produces a provenance paradox: it systematically selects the worst delegates, performing worse than random. We extend the LLM Delegate Protocol (LDP) with delegation contracts that bound authority through explicit objectives, budgets, and failure policies; a claimed-vs-attested identity model that distinguishes self-reported from verified quality; and typed failure semantics enabling automated recovery. In controlled experiments with 10 simulated delegates and validated with real Claude models, routing by self-claimed quality scores performs worse than random selection (simulated: 0.55 vs. 0.68; real models: 8.90 vs. 9.30), while attested routing achieves near-optimal performance (d = 9.51, p < 0.001). Sensitivity analysis across 36 configurations confirms the paradox emerges reliably when dishonest delegates are present. All extensions are backward-compatible with sub-microsecond validation overhead. --- ### A Novel Framework using Intuitionistic Fuzzy Logic with U-Net and U-Net++ Architecture: A case Study of MRI Bain Image Segmentation Authors: Hanuman Verma, Kiho Im, Akshansh Gupta, M. Tanveer Published: 2026-03-15 Categories: eess.IV, cs.LG arXiv: https://arxiv.org/abs/2603.18042v1 Accurate segmentation of brain images from magnetic resonance imaging (MRI) scans plays a pivotal role in brain image analysis and the diagnosis of neurological disorders. Deep learning algorithms, particularly U-Net and U-Net++, are widely used for image segmentation. However, it finds difficult to deal with uncertainty in images. To address this challenge, this work integrates intuitionistic fuzzy logic into U-Net and U-Net++, propose a novel framework, named as IFS U-Net and IFS U-Net++. These models accept input data in an intuitionistic fuzzy representation to manage uncertainty arising from vague ness and imprecise data. This approach effectively handles tissue ambiguity caused by the partial volume effect and boundary uncertainties. To evaluate the effectiveness of IFS U-Net and IFS U-Net++, experiments are conducted on two publicly available MRI brain datasets: the Internet Brain Segmentation Repository (IBSR) and the Open Access Series of Imaging Studies (OASIS). Segmentation performance is quantitatively assessed using Accuracy, Dice Coefficient, and Intersection over Union (IoU). The results demonstrate that the proposed architectures consistently improve segmentation performance by effectively addressing uncertainty --- ### The Voice Behind the Words: Quantifying Intersectional Bias in SpeechLLMs Authors: Shree Harsha Bokkahalli Satish, Christoph Minixhofer, Maria Teleki, James Caverlee, Ondřej Klejch, Peter Bell, Gustav Eje Henter, Éva Székely Published: 2026-03-15 Categories: eess.AS, cs.CL, cs.SD arXiv: https://arxiv.org/abs/2603.16941v2 Speech Large Language Models (SpeechLLMs) process spoken input directly, retaining cues such as accent and perceived gender that were previously removed in cascaded pipelines. This introduces speaker identity dependent variation in responses. We present a large-scale intersectional evaluation of accent and gender bias in three SpeechLLMs using 2,880 controlled interactions across six English accents and two gender presentations, keeping linguistic content constant through voice cloning. Using pointwise LLM-judge ratings, pairwise comparisons, and Best-Worst Scaling with human validation, we detect recurring directional disparities. Eastern European-accented speech receives lower helpfulness scores, particularly for female-presenting voices. Responses remain polite but differ in helpfulness. While LLM judges capture the directional trend of these biases, human evaluators exhibit significantly higher sensitivity, showing stronger accent-level contrasts. --- ### On the Degrees of Freedom of Gridded Control Points in Learning-Based Medical Image Registration Authors: Wen Yan, Qianye Yang, Yipei Wang, Shonit Punwani, Mark Emberton, Vasilis Stavrinides, Yipeng Hu, Dean Barratt Published: 2026-03-15 Categories: eess.IV, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2603.16940v1 Many registration problems are ill-posed in homogeneous or noisy regions, and dense voxel-wise decoders can be unnecessarily high-dimensional. A sparse control-point parameterisation provides a compact, smooth deformation representation while reducing memory and improving stability. This work investigates the required control points for learning-based registration network development. We present GridReg, a learning-based registration framework that replaces dense voxel-wise decoding with displacement predictions at a sparse grid of control points. This design substantially cuts the parameter count and memory while retaining registration accuracy. Multiscale 3D encoder feature maps are flattened into a 1D token sequence with positional encoding to retain spatial context. The model then predicts a sparse gridded deformation field using a cross-attention module. We further introduce grid-adaptive training, enabling an adaptive model to operate at multiple grid sizes at inference without retraining. This work quantitatively demonstrates the benefits of using sparse grids. Using three data sets for registering prostate gland, pelvic organs and neurological structures, the results suggested a significant improvement with the usage of grid-controled displacement field. Alternatively, the superior registration performance was obtained using the proposed approach, with a similar or less computational cost, compared with existing algorithms that predict DDFs or displacements sampled on scattered key points. --- ### Solution for 10th Competition on Ambivalence/Hesitancy (AH) Video Recognition Challenge using Divergence-Based Multimodal Fusion Authors: Aislan Gabriel O. Souza, Agostinho Freire, Leandro Honorato Silva, Igor Lucas B. da Silva, João Vinícius R. de Andrade, Gabriel C. de Albuquerque, Lucas Matheus da S. Oliveira, Mário Stela Guerra, Luciana Machado Published: 2026-03-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.16939v1 We address the Ambivalence/Hesitancy (A/H) Video Recognition Challenge at the 10th ABAW Competition (CVPR 2026). We propose a divergence-based multimodal fusion that explicitly measures cross-modal conflict between visual, audio, and textual channels. Visual features are encoded as Action Units (AUs) extracted via Py-Feat, audio via Wav2Vec 2.0, and text via BERT. Each modality is processed by a BiLSTM with attention pooling and projected into a shared embedding space. The fusion module computes pairwise absolute differences between modality embeddings, directly capturing the incongruence that characterizes A/H. On the BAH dataset, our approach achieves a Macro F1 of 0.6808 on the validation test set, outperforming the challenge baseline of 0.2827. Statistical analysis across 1{,}132 videos confirms that temporal variability of AUs is the dominant visual discriminator of A/H. --- ### Cryptographic Runtime Governance for Autonomous AI Systems: The Aegis Architecture for Verifiable Policy Enforcement Authors: Adam Massimo Mazzocchetti Published: 2026-03-15 Categories: cs.CR, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2603.16938v1 Contemporary AI governance frameworks rely heavily on post hoc oversight, policy guidance, and behavioral alignment techniques, yet these mechanisms become fragile as systems gain autonomy, speed, and operational opacity. This paper presents Aegis, a runtime governance architecture for autonomous AI systems that treats policy and legal constraints as execution conditions rather than advisory principles. Aegis binds each governed agent to a cryptographically sealed Immutable Ethics Policy Layer (IEPL) at system genesis and enforces external emissions through an Ethics Verification Agent (EVA), an Enforcement Kernel Module (EKM), and an Immutable Logging Kernel (ILK). Amendments to the governing policy layer require quorum approval and redeclaration of the system trust root; verified violations trigger autonomous shutdown and generation of auditable proof artifacts. We evaluate the architecture within the Civitas runtime using three operational measures: proof verification latency under tamper conditions, publication overhead, and alignment retention performance relative to an ungoverned baseline. In controlled trials, Aegis demonstrates median proof verification latency of 238 ms, median publication overhead of approximately 9.4 ms, and higher alignment retention than the baseline condition across matched tasks. We argue that these results support a shift in AI governance from discretionary oversight toward verifiable runtime constraint. Rather than claiming to resolve machine ethics in the abstract, the proposed architecture seeks to show that policy violating behavior can be rendered operationally non executable within a controlled runtime governance framework. The paper concludes by discussing methodological limits, evidentiary implications, and the role of proof oriented governance in high assurance AI deployment. --- ### Patient-level validation of a foundation-model pipeline for pediatric lung sounds: physician-annotated adventitious events are recognized, disease groups are not reliably predicted Authors: Izzet Turkalp Akbasli, Oguzhan Serin Published: 2026-03-15 Categories: cs.SD, cs.LG arXiv: https://arxiv.org/abs/2603.15688v2 Background: Lung-sound classifiers are usually evaluated on recording- or event-level splits, although each child contributes many recordings. We evaluated a foundation-model pipeline (PulmoVec) with the patient as the unit of partitioning and asked whether disease group can be predicted beyond age and sex. Methods: We analyzed 19693 physician-annotated respiratory events from 736 children in the public SPRSound database. A frozen Health Acoustic Representations (HeAR) encoder with low-rank adapters was trained for screening (normal versus adventitious), sound pattern (normal, crackles, wheeze/rhonchi) and disease group (pneumonia, bronchial disease, normal/other); event probabilities were stacked with age, sex and auscultation site. The primary analysis was nested patient-grouped cross-validation; comparators were the majority class, annotated event duration and demographics. Results: The area under the receiver operating characteristic curve (AUC) was 0.95 (95% CI 0.94 to 0.96) for both screening and sound pattern, against 0.78 and 0.75 for event duration alone; the positive predictive value for adventitious events was 0.72. Disease-group prediction reached an AUC of 0.58 (95% CI 0.54 to 0.62), and its accuracy was below the majority-class rate at event level (0.44 versus 0.62) and at patient level (0.46 versus 0.60). Conclusions: Recognition of physician-annotated adventitious events holds under patient-level validation and is not explained by event duration or demographics, although its positive predictive value would limit unaided use. Prediction of disease group, a clinical diagnosis, from lung-sound events alone was not demonstrated. Lung-sound studies should partition data by patient and report non-acoustic comparators. --- ### PulmoVec: A Two-Stage Stacking Meta-Learning Architecture Built on the HeAR Foundation Model for Multi-Task Classification of Pediatric Respiratory Sounds Authors: Izzet Turkalp Akbasli, Oguzhan Serin Published: 2026-03-15 Categories: cs.SD, cs.LG arXiv: https://arxiv.org/abs/2603.15688v1 Background: Respiratory diseases are a leading cause of childhood morbidity and mortality, yet lung auscultation remains subjective and limited by inter-listener variability, particularly in pediatric populations. Existing AI approaches are further constrained by small datasets and single-task designs. We developed PulmoVec, a multi-task framework built on the Health Acoustic Representations (HeAR) foundation model for classification of pediatric respiratory sounds. Methods: In this retrospective analysis of the SPRSound database, 24,808 event-level annotated segments from 1,652 pediatric patients were analyzed. Three task-specific classifiers were trained for screening, sound-pattern recognition, and disease-group prediction. Their out-of-fold probability outputs were combined with demographic metadata in a LightGBM stacking meta-model, and event-level predictions were aggregated to the patient level using ensemble voting. Results: At the event level, the screening model achieved an ROC-AUC of 0.96 (95% CI, 0.95-0.97), the sound-pattern recognition model a macro ROC-AUC of 0.96 (95% CI, 0.96-0.97), and the disease-group prediction model a macro ROC-AUC of 0.94 (95% CI, 0.93-0.94). At the patient level, disease-group classification yielded an accuracy of 0.74 (95% CI, 0.71-0.77), a weighted F1-score of 0.73, and a macro ROC-AUC of 0.91 (95% CI, 0.90-0.93). Stacking improved performance across all tasks compared with base models alone. Conclusions: PulmoVec links event-level acoustic phenotyping with patient-level clinical classification, supporting the potential of foundation-model-based digital auscultation in pediatric respiratory medicine. Multi-center external validation across devices and real-world conditions remains essential. --- ### Evidential Domain Adaptation for Remaining Useful Life Prediction with Incomplete Degradation Authors: Yubo Hou, Mohamed Ragab, Yucheng Wang, Min Wu, Abdulla Alseiari, Chee-Keong Kwoh, Xiaoli Li, Zhenghua Chen Published: 2026-03-15 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2603.15687v1 Accurate Remaining Useful Life (RUL) prediction without labeled target domain data is a critical challenge, and domain adaptation (DA) has been widely adopted to address it by transferring knowledge from a labeled source domain to an unlabeled target domain. Despite its success, existing DA methods struggle significantly when faced with incomplete degradation trajectories in the target domain, particularly due to the absence of late degradation stages. This missing data introduces a key extrapolation challenge. When applied to such incomplete RUL prediction tasks, current DA methods encounter two primary limitations. First, most DA approaches primarily focus on global alignment, which can misaligns late degradation stage in the source domain with early degradation stage in the target domain. Second, due to varying operating conditions in RUL prediction, degradation patterns may differ even within the same degradation stage, resulting in different learned features. As a result, even if degradation stages are partially aligned, simple feature matching cannot fully align two domains. To overcome these limitations, we propose a novel evidential adaptation approach called EviAdapt, which leverages evidential learning to enhance domain adaptation. The method first segments the source and target domain data into distinct degradation stages based on degradation rate, enabling stage-wise alignment that ensures samples from corresponding stages are accurately matched. To address the second limitation, we introduce an evidential uncertainty alignment technique that estimates uncertainty using evidential learning and aligns the uncertainty across matched stages. --- ### Life cycle assessment for all organic chemicals Authors: Shaohan Chen, Tim Langhorst, Julian Nöhl, Christopher Oberschelp, Martin Pillich, Johannes Schilling, André Bardow Published: 2026-03-15 Categories: physics.chem-ph, cs.CE, cs.LG arXiv: https://arxiv.org/abs/2603.15686v1 Chemicals are embedded in nearly every aspect of modern society, yet their production poses substantial sustainability concerns. Achieving a sustainable chemical industry requires detailed Life Cycle Assessment (LCA); however, current assessments face many unknowns due to limited, partly inconsistent, and untransparent data coverage since existing Life Cycle Inventory (LCI) databases account for only a tiny fraction of traded chemicals. Here, we introduce the Chemical RetrosYnthesiS for Transparent Assessment of Life-cycles (CRYSTAL) framework, which automatically generates consistent and transparent LCI data for organic chemicals based on their molecular structure using retrosynthesis and machine-learned gate-to-gate inventories. Using the predictive power of CRYSTAL, we create a consistent database for more than 70000 organic chemicals, comprising over 110000 transparent LCI datasets that quantify both feedstock and energy demands, together with associated auxiliary materials, biosphere flows, and waste flows. From this comprehensive database, we identify 50 key environmental hotspots driving high impacts of organic chemical production across multiple environmental categories and pivotal hub chemicals that are most critical for downstream chemical production. In providing this comprehensive data foundation, the CRYSTAL framework offers systematic guidance for targeted engineering and policy interventions. Its transparent, modular nature is designed to shift chemical LCA from a reliance on "unknown unknowns" to a collaboratively improvable mapping of "known unknowns". --- ### DASH: Dynamic Audio-Driven Semantic Chunking for Efficient Omnimodal Token Compression Authors: Bingzhou Li, Tao Huang Published: 2026-03-15 Categories: cs.MM, cs.AI, cs.CV, cs.SD arXiv: https://arxiv.org/abs/2603.15685v2 Omnimodal large language models (OmniLLMs) jointly process audio and visual streams, but the resulting long multimodal token sequences make inference prohibitively expensive. Existing compression methods typically rely on fixed window partitioning and attention-based pruning, which overlook the piecewise semantic structure of audio-visual signals and become fragile under aggressive token reduction. We propose Dynamic Audio-driven Semantic cHunking (DASH), a training-free framework that aligns token compression with semantic structure. DASH treats audio embeddings as a semantic anchor and detects boundary candidates via cosine-similarity discontinuities, inducing dynamic, variable-length segments that approximate the underlying piecewise-coherent organization of the sequence. These boundaries are projected onto video tokens as a soft temporally co-registered segmentation prior. Within each segment, token retention is determined by a tri-signal importance estimator that fuses structural boundary cues, representational distinctiveness, and attention-based salience, mitigating the sparsity bias of attention-only selection. This structure-aware allocation preserves transition-critical tokens while reducing redundant regions. Extensive experiments on AVUT, VideoMME, and WorldSense demonstrate that DASH maintains competitive or superior accuracy while achieving higher compression ratios compared to prior methods. Code is available at: https://github.com/laychou666/DASH. --- ### State-Dependent Safety Failures in Multi-Turn Language Model Interaction Authors: Pengcheng Li, Jie Zhang, Tianwei Zhang, Han Qiu, Zhang kejun, Weiming Zhang, Nenghai Yu, Wenbo Zhou Published: 2026-03-15 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2603.15684v2 Safety alignment in large language models is typically evaluated under isolated queries, yet real-world use is inherently multi-turn. Although multi-turn jailbreaks are empirically effective, the structure of conversational safety failure remains insufficiently understood. In this work, we study safety failures from a state-space perspective and show that many multi-turn safety failures in current safety-aligned language models arise from contextual state evolution, a regime that is not fully captured by isolated prompt-level analyses alone. We introduce STAR, a state-oriented diagnostic framework that treats dialogue history as a state transition operator and enables controlled analysis of safety behavior along interaction trajectories. Rather than optimizing attack strength, STAR provides a principled probe of how aligned models traverse the safety boundary under autoregressive conditioning. Across multiple frontier language models, we find that systems that appear robust under static evaluation can undergo rapid and reproducible safety collapse under structured multi-turn interaction. Mechanistic analysis reveals monotonic drift away from refusal-related representations and abrupt phase transitions induced by role-conditioned context. Together, these findings motivate viewing language model safety as a dynamic, state-dependent process defined over conversational trajectories. --- ### Beyond Distance: Quantifying Point Cloud Dynamics with Persistent Homology and Dynamic Optimal Transport Authors: Yixin Wang, Ting Gao, Jinqiao Duan Published: 2026-03-15 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2603.15683v1 We introduce a framework for analyzing topological tipping in time-evolutionary point clouds by extending the recently proposed Topological Optimal Transport (TpOT) distance. While TpOT unifies geometric, homological, and higher-order relations into one metric, its global scalar distance can obscure transient, localized structural reorganizations during dynamic phase transitions. To overcome this limitation, we present a hierarchical dynamic evaluation framework driven by a novel topological and hypergraph reconstruction strategy. Instead of directly interpolating abstract network parameters, our method interpolates the underlying spatial geometry and rigorously recomputes the valid topological structures, ensuring physical fidelity. Along this geodesic, we introduce a set of multi-scale indicators: macroscopic metrics (Topological Distortion and Persistence Entropy) to capture global shifts, and a novel mesoscopic dual-perspective Hypergraph Entropy (node-perspective and edge-perspective) to detect highly sensitive, asynchronous local rewirings. We further propagate the cycle-level entropy change onto individual vertices to form a point-level topological field. Extensive evaluations on physical dynamical systems (Rayleigh-Van der Pol limit cycles, Double-Well cluster fusion), high-dimensional biological aggregation (D'Orsogna model), and longitudinal stroke fMRI data demonstrate the utility of combining transport-based alignment with multi-scale entropy diagnostics for dynamic topological analysis. --- ### Flood Risk Follows Valleys, Not Grids: Graph Neural Networks for Flash Flood Susceptibility Mapping in Himachal Pradesh with Conformal Uncertainty Quantification Authors: Paras Sharma, Swastika Sharma Published: 2026-03-15 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.15681v1 Flash floods are the most destructive natural hazard in Himachal Pradesh (HP), India, causing over 400 fatalities and $1.2 billion in losses in the 2023 monsoon season alone. Existing risk maps treat every pixel independently, ignoring the basic fact that flooding upstream raises risk downstream. We address this with a Graph Neural Network (GraphSAGE) trained on a watershed connectivity graph (460 sub-watersheds, 1,700 directed edges), built from a six-year Sentinel-1 SAR flood inventory (2018-2023, 3,000 events) and 12 environmental variables at 30 m resolution. Four pixel-based ML models (RF, XGBoost, LightGBM, stacking ensemble) serve as baselines. All models are evaluated with leave-one-basin-out spatial cross-validation to avoid the 5-15% AUC inflation of random splits. Conformal prediction produces the first HP susceptibility maps with statistically guaranteed 90% coverage intervals. The GNN achieved AUC = 0.978 +/- 0.017, outperforming the best baseline (AUC = 0.881) and the published HP benchmark (AUC = 0.88). The +0.097 gain confirms that river connectivity carries predictive signal that pixel-based models miss. High-susceptibility zones overlap 1,457 km of highways (including 217 km of the Manali-Leh corridor), 2,759 bridges, and 4 major hydroelectric installations. Conformal intervals achieved 82.9% empirical coverage on the held-out 2023 test set; lower coverage in high-risk zones (45-59%) points to SAR label noise as a target for future work. --- ### Computational Analysis of Semantic Connections Between Herman Melville Reading and Writing Authors: Nudrat Habib, Elisa Barney Smith, Steven Olsen Smith Published: 2026-03-15 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.14674v1 This study investigates the potential influence of Herman Melville reading on his own writings through computational semantic similarity analysis. Using documented records of books known to have been owned or read by Melville, we compare selected passages from his works with texts from his library. The methodology involves segmenting texts at both sentence level and non-overlapping 5-gram level, followed by similarity computation using BERTScore. Rather than applying fixed thresholds to determine reuse, we interpret precision, recall, and F1 scores as indicators of possible semantic alignment that may suggest literary influence. Experimental results demonstrate that the approach successfully captures expert-identified instances of similarity and highlights additional passages warranting further qualitative examination. The findings suggest that semantic similarity methods provide a useful computational framework for supporting source and influence studies in literary scholarship. --- ### A Single-Sample Polylogarithmic Regret Bound for Nonstationary Online Linear Programming Authors: Haoran Xu, Owen Shen, Peter Glynn, Yinyu Ye, Patrick Jaillet Published: 2026-03-15 Categories: cs.DS, cs.LG arXiv: https://arxiv.org/abs/2603.14673v1 We study nonstationary Online Linear Programming (OLP), where $n$ orders arrive sequentially with reward-resource consumption pairs that form a sequence of independent, but not necessarily identically distributed, random vectors. At the beginning of the planning horizon, the decision-maker is provided with a resource endowment that is sufficient to fulfill a significant portion of the requests. The decision-maker seeks to maximize the expected total reward by making immediate and irrevocable acceptance or rejection decisions for each order, subject to this resource endowment. We focus on the challenging single-sample setting, where only one sample from each of the $n$ distributions is available at the start of the planning horizon. We propose a novel re-solving algorithm that integrates a dynamic programming perspective with the dual-based frameworks traditionally employed in stationary environments. In the large-resource regime, where the resource endowment scales linearly with the number of orders, we prove that our algorithm achieves $O((\log n)^2)$ regret across a broad class of nonstationary distribution sequences. Our results demonstrate that polylogarithmic regret is attainable even under significant environmental shifts and minimal data availability, bridging the gap between stationary OLP and more volatile real-world resource allocation problems. --- ### Seamless Deception: Larger Language Models Are Better Knowledge Concealers Authors: Dhananjay Ashok, Ruth-Ann Armstrong, Jonathan May Published: 2026-03-15 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2603.14672v2 Language Models (LMs) may acquire harmful knowledge, and yet feign ignorance of these topics when under audit. Inspired by the recent discovery of deception-related behaviour patterns in LMs, we aim to train classifiers that detect when a LM is actively concealing knowledge. Initial findings on smaller models show that classifiers can detect concealment more reliably than human evaluators, with gradient-based concealment proving easier to identify than prompt-based methods. However, contrary to prior work, we find that the classifiers do not reliably generalize to unseen model architectures and topics of hidden knowledge. Most concerningly, the identifiable traces associated with concealment become fainter as the models increase in scale, with the classifiers achieving no better than random performance on any model exceeding 70 billion parameters. Our results expose a key limitation in black-box-only auditing of LMs and highlight the need to develop robust methods to detect models that are actively hiding the knowledge they contain. ---