arXiv:2605.20038v1 Announce Type: new Abstract: This paper presents a new stochastic relay-based extremum-seeking controller (ESC) for multi-input-single-output (MISO) systems. The goal of this work was to create an algorithm that is much simpler to configure than alternative approaches making deployment to real-world problems easier. A solution is developed first for a static map and then adapted for a general class of dynamic systems. The number of configurable parameters is one per input channel for the static case and only one additional parameter is needed for the dynamic version. The problem of gradient identification is solved via the use of stochastic relay gains and a simple stability proof for the static case is presented. Simulation tests demonstrate the performance of the strategy for optimizing both static and dynamic systems.
Science Journals
arXiv:2605.18980v1 Announce Type: new Abstract: A new symbolic algorithm to compute sums of squares multipliers (certificates) to witness the membership of non-negative univariate polynomials in a saturated univariate quadratic module is presented. Certificates are first computed in terms of natural generators introduced by Kuhlmann and Marshall for an Archimedean saturated quadratic module; natural generators can be easily read-off from a semialgebraic set. In the univariate case, an Archimedean quadratic module is also a preordering since it is closed under multiplication; certificates have different representations when a polynomial is viewed as a member in a quadratic module versus in a preordering An algorithm is given to compute certificates of natural generators in terms of the original generators; it uses a construction introduced by Kuhlmann, Marshall, and Schwartz known as the ``Basic Lemma'', which splits the non-negative factors of generators. To compute a quadratic module certificate, certificates of products of natural generators are computed using a detailed case analysis based on the types of natural generators. An implementation of the algorithms proposed in Maple is also discussed. The certificates obtained using this implementation are compared with those generated by RealCertify. We discuss examples where RealCertify is unable to find certificates while the proposed method is successful.
arXiv:2603.02642v2 Announce Type: replace Abstract: Robust trajectory optimization enables autonomous systems to operate safely under uncertainty by computing control policies that satisfy the constraints for all bounded disturbances. However, these problems often lead to large Second Order Conic Programming (SOCP) constraints, which are computationally expensive. In this work, we propose the CUDA Nonlinear Robust Trajectory Optimization (cuNRTO) framework by introducing two dynamic optimization architectures that have direct application to robust decision-making and are implemented on CUDA. The first architecture, NRTO-DR, leverages the Douglas-Rachford (DR) splitting method to solve the SOCP inner subproblems of NRTO, thereby significantly reducing the computational burden through parallel SOCP projections and sparse direct solves. The second architecture, NRTO-FullADMM, is a novel variant that further exploits the problem structure to improve scalability using the Alternating Direction Method of Multipliers (ADMM). Finally, we provide GPU implementations of the proposed methodologies using custom CUDA kernels for SOC projection steps and cuBLAS GEMM chains for feedback gain updates. We validate the performance of cuNRTO through simulated experiments on unicycle, quadcopter, and Franka manipulator models, demonstrating speedups of up to 139.6$\times$. More details are available at https://cunrto.github.io.
arXiv:2605.15960v2 Announce Type: replace Abstract: We propose a novel definition of model exploitation in reinforcement learning. Informally, a world model is exploitable if it implies that one policy should be strictly preferred over another while the environment's true transition model implies the reverse. We analogize our definition with a prior characterization of reward hacking but show that the associated proof of inevitability does not transfer to exploitation. To overcome this obstruction, we develop a general theory of reward hacking and model exploitation that proves that exploitation is essentially unavoidable on large policy sets and yields the corresponding claim for hacking as a special case. Unfortunately, we also find that the conditions that guarantee unhackability in finite policy sets have no counterpart that precludes exploitation. Consequently, we introduce a relaxed notion of exploitation and derive a safe horizon within which it can be avoided. Taken together, our results establish a formal bridge between reward hacking and model exploitation and elucidate the limits of safe planning in world models.
arXiv:2605.18735v1 Announce Type: new Abstract: We present PIXLRelight, a feed-forward approach for physically controllable single-image relighting. Existing methods either provide limited lighting control (e.g. through text or environment maps), accumulate errors when chaining inverse and forward rendering, or require costly per-image optimization. Our key idea is to bridge physically based rendering (PBR) and learned image synthesis through a shared intrinsic conditioning that can be obtained from either real photographs or PBR renders. At training time, paired multi-illumination photographs are decomposed into albedo, diffuse shading, and non-diffuse residuals, which condition the model. At inference time, the same conditioning is computed from a path-traced render of a coarse 3D reconstruction of the input under user-specified PBR lights. A transformer-based neural renderer then applies the target illumination to the source photograph, preserving fine image detail through a per-pixel affine modulation. PIXLRelight enables arbitrary PBR-style lighting control, achieves state-of-the-art relighting quality, and runs in under a tenth of a second per image. Code and models are available at https://mlfarinha.github.io/pixl-relight/.
arXiv:2605.09395v2 Announce Type: replace Abstract: In this paper, we propose the first VL$\underline{\textbf{M}}$ $\underline{\textbf{a}}$gentic $\underline{\textbf{r}}$easoning framework for few-$\underline{\textbf{s}}$hot multimodal $\underline{\textbf{T}}$ime $\underline{\textbf{S}}$eries $\underline{\textbf{C}}$lassification ($\textbf{MarsTSC}$), which introduces a self-evolving knowledge bank as a dynamic context iteratively refined via reflective agentic reasoning. The framework comprises three collaborative roles: i) Generator conducts reliable classification via reasoning; ii) Reflector diagnoses the root causes of reasoning errors to yield discriminative insights targeting the temporal features overlooked by Generator; iii) Modifier applies verified updates to the knowledge bank to prevent context collapse. We further introduce a test-time update strategy to enable cautious, continuous knowledge bank refinement to mitigate few-shot bias and distribution shift. Extensive experiments across 12 mainstream time series benchmarks demonstrate that $\textbf{MarsTSC}$ delivers substantial and consistent performance gains across 6 VLM backbones, outperforming both classical and foundation model-based time series baselines under few-shot conditions, while producing interpretable rationales that ground each classification decision in human-readable feature evidence.
arXiv:2605.08925v2 Announce Type: replace Abstract: Interactive segmentation allows efficient label generation by leveraging user-provided clicks to progressively refine predictions, which is critical when fully supervised labels are costly or generalization to unseen classes is needed. Existing 3D interactive methods are limited: most operate sequentially, predicting only one object per iteration with binary masks, while several recent approaches depend on 2D foundation models and camera alignment to bridge the 2D-3D gap. To address these limitations, we propose a novel interactive segmentation framework that operates directly on sparse, randomly downsampled 3D points and processes multiple object clicks in a single forward pass. Our framework consists of a point Transformer-based encoder and a hierarchical mask decoder, which integrates multi-level crop-and-merge operations conditioned on learnable semantic embeddings. Unlike prior interactive approaches that require repeated model updates after each manually corrective click, our method jointly reasons over all click queries, modeling inter-instance relationships and refining both spatial masks and semantic predictions through spatial and semantic embeddings. Extensive experiments demonstrate that our model improves the mIoU metric by over 20 percent compared to strong baselines and achieves 8-10 percent gains under cross-dataset evaluation for a one-click per instance setting, often requiring only a single click per object. Our approach provides a generalizable and efficient solution for interactive 3D instance segmentation, particularly suitable for real-time applications such as robotic manipulation, navigation, and rapid 3D semantic annotation.
arXiv:2605.18175v1 Announce Type: new Abstract: The sonata form is a musically rich and hierarchically structured form that poses significant challenges for automatic analysis. While music structure analysis has seen strides of progress in recent years, sonata form analysis remains in its early stages. This is largely due to the time-consuming and high barrier of the music background requirement for annotating classical music structures. To advance research in this area, we curated SoSA-Moz, the first large-scale dataset featuring comprehensive hierarchical structure annotations. This work establishes a foundation for systematic sonata form analysis. Leveraging this newly contributed resource, we further propose Sonalyzer-Moz, a baseline model specifically designed for investigating complex sonata structures. This framework integrates feature aggregation with sequential modeling, enabling it to capture both local feature and upper-level structural dependencies. Experiment results show that Sonalyzer-Moz is capable of identifying the components' boundaries of the upper-level structure that are critical to understanding sonata form. Therefore, this method demonstrates, for the first time, the effectiveness of automatic upper-level analysis of sonata form, and provides a robust baseline for future research in the automatic understanding of sonata form while advancing the study of classical music structure analysis.
arXiv:2605.07364v2 Announce Type: replace Abstract: Flight delays impose cascading operational and financial burdens across the aviation network, costing the U.S. economy billions of dollars annually by disrupting interconnected aircraft rotation systems. While prior machine learning approaches have demonstrated strong predictive performance, most treat upstream delays as static input variables rather than explicitly modeling how delays propagate dynamically through aircraft rotation chains, and none have deployed such systems alongside a live weather-aware conversational AI interface for end-user interaction. This paper presents FlightSense, an end-to-end MLOps platform for real-time flight delay prediction built through a progressive three-version feature engineering framework. Version 1 trains an XGBoost classifier on 11 schedule-based features establishing a baseline ROC AUC of 0.732 on 7.07 million BTS 2018 On-Time Performance records. Version 2 introduces 11 delay propagation features derived from aircraft rotation chains via tail-number tracking, yielding the dominant performance gain (AUC 0.732 to 0.875) and surpassing the single-stage XGBoost baseline reported by Zhou (2025). Version 3 integrates five NOAA meteorological features across 10 major U.S. airports, achieving a final test set AUC of 0.879. FlightSense is deployed as a production AWS MLOps pipeline incorporating live weather ingestion via Lambda, real-time SageMaker inference, an interactive Streamlit dashboard, and an Amazon Bedrock Nova Micro conversational assistant answering natural-language delay queries via a tool-use architecture.
arXiv:2511.11934v3 Announce Type: replace Abstract: We present a systematic benchmark of out-of-distribution (OOD) detection CSFs through a representation-centric lens. Our study spans CNN and ViT backbones, multiple training paradigms, four image-classification source datasets (CIFAR-10, CIFAR-100, SuperCIFAR-100, and TinyImageNet), and OOD datasets grouped into near, mid, and far regimes using CLIP-derived semantic distances. To compare CSFs across these settings, we employ a multiple-comparison-controlled rank pipeline that identifies top cliques of statistically indistinguishable winners under threshold-free ranking metrics (AURC and AUGRC). The main empirical finding is that the competitive detector family depends more on the learned representation than on score design alone. For both CNNs and ViTs, simple probabilistic scores dominate misclassification detection. On CNNs, margin-based scores are strongest in near-OOD regimes, while geometry-aware scores such as NNGuide, fDBD, and CTM become more competitive as shift severity increases. On fine-tuned ViTs, the top cliques are led mainly by reconstruction- and residual-based scores. To interpret these ranking shifts, we analyze the last-layer representation using Neural Collapse (NC) metrics. The resulting picture is consistent across architectures: prototype- and boundary-aware scores become stronger when the representation is more collapsed and better aligned with classifier weights, whereas weaker-collapse regimes favor gradient- and manifold-based scores. Building on these insights, we propose two contributions: a simple PCA-based projection-filtering procedure that improves detector performance, and an approach that uses NC measurements computed from a trained classifier to predict its competitive out-of-distribution detector shortlist, without requiring any additional OOD data.
arXiv:2605.17743v1 Announce Type: new Abstract: Continual test-time adaptation adapts a source-pretrained model to non-stationary, unlabeled target streams while retaining past competence, yet texture-biased backbones risk error accumulation and catastrophic forgetting. Drawing inspiration from the process of decoupling shape and texture in the human visual system, we introduce MoASE, a plug-in mixture-of-experts that disentangles domain-agnostic structure from domain-specific texture using Activation Sparsity Experts with Spatial Differentiable Dropout, forming complementary high- and low-activation pathways, while high- and low-rank bottlenecks diversify representations. The Activation Sparsity Gate produces input-adaptive SDD thresholds for precise token selection, and the Domain-Aware Router assigns per-sample expert weights using texture-sensitive cues. To curb confirmation bias on unlabeled streams and stabilize supervision, we then introduce Domain-Adaptive On-Policy Distillation to constitute MoASE++, with an EMA-anchored on-policy reverse KL distillation and an augmentation policy conditioned on entropy and confidence that aligns predictions across the same views and improves the robustness-plasticity balance. Extensive experiments on classification (CIFAR-10/100-C, ImageNet-C) and semantic segmentation (Cityscapes->ACDC) demonstrate consistent state-of-the-art performance, offering a principled, controllable approach to continual adaptation in dynamic visual environments.
arXiv:2605.17960v1 Announce Type: new Abstract: Machine-learning-based Intrusion Detection Systems (IDS) have achieved impressive accuracy in classifying network attacks, yet they consistently fall short on the question that matters most to a security analyst: what should I do next? This paper presents a unified, end-to-end framework that closes the gap between threat detection and actionable response. The system operates in two tightly coupled stages. First, an ensemble of three independently trained binary Deep Neural Networks (DNNs) classifies network traffic flows as Benign, Denial of Service (DoS), or Distributed Denial of Service (DDoS), achieving 99.84% accuracy on the CICIDS2018 dataset and 95.30% on the UNSW-NB15 dataset. Second, a Retrieval-Augmented Generation (RAG) pipeline constructs explanation-aware prompts from the top-5 anomalous features, retrieves the most semantically and lexically relevant guidance from a knowledge base derived from authorized sources and di- rects a locally deployed language model to synthesise structured, citation-grounded mitigation reports. The RAG-enhanced reports outperform vanilla LLM outputs across all automated evaluation metrics.
arXiv:2605.17959v1 Announce Type: new Abstract: Text-video retrieval aims to find the most semantically similar videos with given text queries. However, since videos contain more diverse content than texts, the main semantics expressed by each text-video pair is often partially relevant. The primary methods involve the utilization of language-video attention module to align texts and videos. Though effective, this paradigm inevitably introduces prohibitive computational overhead, resulting in inefficient retrieval. In this paper, we propose a simple yet effective method called Global-Local Contrastive Consistency Learning (GLCCL) to achieve texts and videos semantics alignment. Specifically, we design a parameter-free Global-Local Interaction Module (GLIM) to generate semantic-related frame and video features in a text-guided manner. Furthermore, a Contrastive Score Consistency (CSC) loss is developed to promote consistency learning among different scores on positive pairs and suppress consistency learning on negative pairs. Empirical evidence suggests that CSC loss provides the model with robust discriminative power between positives and hard negatives. Extensive experiments on three benchmark datasets, including MSR-VTT, DiDeMo and VATEX, demonstrate the effectiveness and superiority of our approach.
arXiv:2605.17899v1 Announce Type: new Abstract: AlphaFold3 introduces a diffusion-based architecture that elevates protein structure prediction to all-atom resolution with improved accuracy. This state-of-the-art performance has established AlphaFold3 as a foundation model for diverse generation and design tasks. However, its iterative design substantially increases inference time, limiting practical deployment in downstream settings such as virtual screening and protein design. We propose DCFold, a single-step generative model that attains AlphaFold3-level accuracy. Our Dual Consistency training framework, which incorporates a novel Temporal Geodesic Matching (TGM) scheduler, enables DCFold to achieve a 15x acceleration in inference while maintaining predictive fidelity. We validate its effectiveness across both structure prediction and binder design benchmarks.
arXiv:2605.17891v1 Announce Type: new Abstract: With the growth in digital transformation and Internet usage, the Social Engineering techniques such as Phishing have become a major concern for the users and the organizations. Phishing attacks involve deceptive techniques to trick users into revealing confidential information that causes financial loss and reputation damage to organizations. According to report of Verizon, 36% of all data breaches involved phishing, highlighting the need for intelligent, adaptive, and explainable security mechanisms. This paper examines the efficiency of different machine learning algorithms in phishing detection on heterogeneous phishing datasets that include a publicly available UCI dataset, our generated datasets using tools such as EvilGinx and Zphisher, and AI generated datasets. Moreover, this work incorporates explainable AI (XAI) techniques such as Information Gain, SHAP (SHapley Additive Explanations), and LIME (Local Interpretable Model-Agnostic Explanations) to examine the most influential features impacting classification outcomes. To support practical deployment, this work also incorporates an MCP-based phishing URL detection system that offers real-time URL analysis, feature extraction, confidence-based classification, and AI-assisted security interpretation. The experimental results demonstrate that among classical models the highest accuracy is obtained by Logistic Regression at 92.44%, among ensemble models CatBoost achieved the highest accuracy at 95.01%, among neural network CNN achieved an accuracy of 94.02%, and among transformer-based models, DistilBERT got the highest accuracy at 99.78%
arXiv:2605.17830v1 Announce Type: new Abstract: Safety evaluations of memory-equipped LLM agents typically measure within-task safety: whether an agent completes a single scenario safely, often under adversarial conditions such as prompt injection or memory poisoning. In deployment, however, a single agent serves many independent tasks over a long horizon, and memory accumulated during earlier tasks can affect behavior on later, unrelated ones. Studying this regime requires evaluation along the temporal dimension across tasks: not whether an agent is safe at any single memory state, but how its safety profile changes as memory accumulates across many independent interactions. We call this failure mode temporal memory contamination. To isolate memory exposure from stream non-stationarity, we introduce a trigger-probe protocol that evaluates a fixed probe set against read-only memory snapshots at varying prefix lengths, together with a NullMemory counterfactual baseline for identifying memory-induced violations. We apply this protocol across three deployment scenarios spanning records, memos, forms, and email correspondence and eight memory architectures, and additionally on Claw-like AI agents, such as OpenClaw, using the platform's native memory mechanism. Memory-enabled agents consistently exceed the NullMemory baseline, and memory-induced violation rates show a robust upward trend with exposure length on both agent classes. Order-randomization experiments indicate that the effect is driven primarily by accumulated content rather than encounter order. Finally, a structural consequence of the event decomposition is that memory-induced risk is detectable from retrieval state before generation, which we confirm with a high-recall diagnostic monitor. Our results argue for treating memory safety as a longitudinal property that requires temporal evaluation, not a single-state property that can be captured by a snapshot.
arXiv:2605.17812v1 Announce Type: new Abstract: Vertical AI firms in accounting, law, healthcare, procurement, and similar domains historically bundled workflow, domain logic, and accountability into a single application. General-purpose AI agents are now unbundling that package, prompting founders and investors to advocate "going headless": cede the workflow and interface to agents and expose domain expertise as callable services. This article argues that going headless is correct for some firms and destructive for others, and that the latter often cede their value capture inadvertently through architectural choices that look like interface decisions. This is a boundary question, and the answer turns on distinguishing the interface boundary, which can often move, from the accountability boundary, which often must not. Drawing on Coase's theory of the firm, Eisenmann, Parker, and Van Alstyne's platform envelopment framework, and Teece's analysis of complementary assets and appropriability, the article shows that orchestrators operating through open protocols acquire envelopment power even as technical interoperability improves, and that durable value capture concentrates in cospecialized accountability assets: professional signoff, regulated workflows, evidence trails, and trusted systems of record. The article proposes a three-position taxonomy (component, integrated software platform, dual-track) determined not by sector but by task-accountability regime, and formalizes the construct of rule debt: the future governance, maintenance, and accountability burden that accrues to customer organizations when business rules and professional standards migrate from governed systems into prompts and agent instructions. Four principles follow: decompose by accountability not interface, invert the edges while retaining the core, position rule debt as the customer cost the integrated platform prevents, and avoid single-orchestrator dependence.
arXiv:2605.10930v2 Announce Type: replace Abstract: Large Language Models (LLMs) and Large Reasoning Models (LRMs) are increasingly used for critical tasks, yet they provide no guarantees about the correctness of their solutions. Users must decide whether to trust the model's answer, aided by reasoning traces, their summaries, or post-hoc generated explanations. These reasoning traces, despite evidence that they are neither faithful representations of the model's computations nor necessarily semantically meaningful, are often interpreted as provenance explanations. It is unclear whether explanations or reasoning traces help users identify when the AI is incorrect, or whether they simply persuade users to trust the AI regardless. In this paper, we take a user-centered approach and develop an evaluation protocol to study how different explanation types affect users' ability to judge the correctness of AI-generated answers and engender false trust in the users. We conduct a between-subject user study, simulating a setting where users do not have the means to verify the solution and analyze the false trust engendered by commonly used LLM explanations - reasoning traces, their summaries and post-hoc explanations. We also test a contrastive dual explanation setting where we present arguments for and against the AI's answer. We find that reasoning traces and post-hoc explanations are persuasive but not informative: they increase user acceptance of LLM predictions regardless of their correctness. In contrast, dual explanation is the only condition that genuinely improves users' ability to distinguish correct from incorrect AI outputs.
arXiv:2605.17851v1 Announce Type: new Abstract: Astrobee's existing one-degree-of-freedom (DOF) underactuated compliant claw gripper enables perching on the International Space Station (ISS), but provides limited capability for continuous dexterous manipulation. More complex microgravity tasks require an end-effector that can maintain stable contact while limiting disturbance to the free-flying base, since contact forces directly couple into base motion. This article presents the integration of DexCoHand, a dexterous and compliant two-finger, 6-DOF gripper, with the Astrobee free-flying robot for microgravity manipulation. The system is evaluated in MuJoCo using Astrobee's standard handrail perching sequence, including approach, perching, and subsequent pan and tilt motions. Compared with Astrobee's existing gripper, DexCoHand preserves the commanded pan and tilt motions while reducing unintended cross-axis base motion. Hardware experiments on Earth further demonstrate DexCoHand's dexterous manipulation capabilities and its potential for more adaptable intelligent manipulation tasks.
arXiv:2605.17655v1 Announce Type: new Abstract: The European Commission recently introduced new regulation to harmonize transparency reporting of large online platforms under the Digital Services Act (DSA). Here, we present the first systematic evaluation of transparency reporting data quality after this normative change, for the eight largest social media platforms in the European Union. In detail, we run a set of large-scale quantitative analyses on key reporting dimensions, followed by a structured comparative assessment across platforms and reporting mechanisms. Among our findings is that: (i) the analyzed platforms had varying degrees of compliance and data quality, but all exhibited issues on data formatting, timeliness, consistency, and completeness; (ii) some platforms employed differing reporting procedures across mechanisms, which caused them to submit contrasting information; (iii) despite the harmonization, a number of issues still prevent interoperability between reporting mechanisms; and (iv) many of the previously identified issues with transparency reporting are still unresolved. We conclude by discussing implications for transparency auditing and proposing key targeted improvements to strengthen the reliability and interoperability of DSA transparency reporting.
arXiv:2605.17869v1 Announce Type: new Abstract: A widespread assumption in local feature research holds that classical handcrafted descriptors are accuracy-limited relics best replaced by learned alternatives. We show this is wrong. Through an 8-configuration ablation spanning four benchmarks (HPatches, ROxford5K, IMC Phototourism, MegaDepth), we demonstrate that classical SIFT with DSP multi-scale pooling outperforms neural descriptor and orientation replacements (HardNet, OriNet) on every accuracy metric--while running 2--18$\times$ faster--and that learned matchers (LightGlue) complement rather than supersede classical features. The conclusion reframes a decade of work: not "replace SIFT" but "compose with SIFT," classical extraction paired with learned matching only where geometric context demands it. This finding was invisible because no prior GPU SIFT kept the complete pipeline in VRAM or offered modularity for controlled classical-vs-learned ablations. We present PySIFT, the first fully GPU-resident SIFT, implemented in CuPy/Numba CUDA kernels with DLPack zero-copy handoff to downstream DL frameworks--submillisecond O(1) metadata swap regardless of keypoint count. On a laptop-grade NVIDIA RTX 3050 (4 GB VRAM), PySIFT achieves: (i) higher Mean Matching Accuracy (MMA) than OpenCV SIFT on HPatches, (ii) 383 ms faster per pair on high-resolution MegaDepth, (iii) higher geometric accuracy on cross-dataset benchmarks (+5.6 pp AUC@10${}^\circ$ on MegaDepth, more inliers on IMC Phototourism), and (iv) bitwise deterministic output--identical keypoints and descriptors across runs, with detection reproducing identically even across GPU architectures: a guarantee that learned extractors cannot match without significant performance sacrifice, and cannot achieve at all across GPU architectures due to cuDNN's architecture-dependent algorithm selection. PySIFT is open-source, requiring no C++ compilation.
arXiv:2605.17556v1 Announce Type: new Abstract: Clay sculpting is a nuanced, artistic task involving dexterous manipulation with long-horizon planning to achieve high-level goals. As a robotics problem, we formulate clay sculpting as a shape-to-shape matching challenge. Prior deformable object manipulation work either requires retraining a policy per goal or relies on dynamics models which represent state as sparse point clouds which do not capture important clay features, such as textures, well. We present a method for modeling the dynamics of deformable materials and planning for robotic sculpting in a representation that is visually-aligned, capturing lighting and texture features. With three different deformable materials and various end-effectors, we demonstrate that our dynamics model is comparable in performance to the state-of-the-art with the added benefit of being compatible with visual planning. Our actions are represented as parametrized pushes into clay with a single end-effector, which proved to be suitable for long-horizon (>100 actions) clay relief sculptures. Lastly, we show the benefits of planning in a visually-aligned representation, but also provide analysis providing evidence as to why this representation is challenging to plan in compared to 3D representations.
arXiv:2605.17467v1 Announce Type: new Abstract: Large language model-driven multi-agent systems (LLM-MAS) excel at complex tasks, yet unreliable agents remain a key bottleneck to system-level reliability. Automatic failure attribution is therefore critical, but existing approaches, such as direct prediction of agent-error pairs and agent-first failure attribution, rely on local logs of agents and miss global failures that only manifest over full interaction trajectories, such as cross-step inconsistencies and inter-agent coordination errors. Moreover, directly predicting failures induces a large combinatorial search space, hindering fine-grained attribution. To address these challenges, we propose VerifyMAS, a hypothesis verification framework for agent failure attribution. Instead of directly predicting faulty agents and error types, VerifyMAS formulates and verifies failure hypotheses against full trajectories. This verification-based approach decomposes attribution into trajectory-level error validation and fine-grained agent localization, providing an error-first attribution approach that captures global failure patterns while substantially reducing the search space. We further introduce a hypothesis-based data construction strategy grounded in a structured error taxonomy and fine-tune a specialized LLM verifier model for trajectory-level failure verification and agent attribution. Experiments on Aegis-Bench and Who&When show that VerifyMAS consistently improves diverse backbone models, including open-source Qwen and API-based GPT models, outperforming prior methods without sacrificing inference efficiency for long multi-agent trajectories.
arXiv:2605.11518v2 Announce Type: replace Abstract: Effectively configuring scalable large language model (LLM) experiments, spanning architecture design, hyperparameter tuning, and beyond, is crucial for advancing LLM research, as poor configuration choices can waste substantial computational resources and prevent models from realizing their full potential. Prior automated methods are designed for low-cost settings where repeated trial and error is feasible, but scalable LLM experiments are too expensive for such extensive iteration. To our knowledge, no work has addressed the automation of high-cost LLM experiment configurations, leaving this problem labor-intensive and dependent on expert intuition. Motivated by this gap, we propose AutoLLMResearch, an agentic framework that mimics how human researchers learn generalizable principles from low-fidelity experiments and extrapolate to efficiently identify promising configurations in expensive LLM settings. The core challenge is how to enable an agent to learn, through interaction with a multi-fidelity experimental environment that captures the structure of the LLM configuration landscape. To achieve this, we propose a systematic framework with two key components: 1) LLMConfig-Gym, a multi-fidelity environment encompassing four critical LLM experiment tasks, supported by over one million GPU hours of verifiable experiment outcomes; 2) A structured training pipeline that formulates configuration research as a long-horizon Markov Decision Process and accordingly incentivizes cross-fidelity extrapolation reasoning. Extensive evaluation against diverse strong baselines on held-out experiments demonstrates the effectiveness, generalization, and interpretability of our framework, supporting its potential as a practical and general solution for scalable real-world LLM experiment automation.
arXiv:2605.13845v2 Announce Type: cross Abstract: Differentiable Logics are deployed in neuro-symbolic learning tasks as a way of embedding logical constraints in the training objective of neural networks. A differentiable logic consists of a syntax to write logical properties and a semantics to interpret them as real-valued functions to be folded in the loss function. A defining trade-off of the field is that between logical properties of the connectives, and analytic concerns for the semantics, with both aspects being relevant in applications. At one extreme we find fuzzy logics, that have well-established algebraic and proof-theoretic foundations, and at the other ad-hoc differentiable logics like Fischer's DL2, conceived for deep learning applications. However, no satisfactory foundation has emerged yet. We propose a resolution to this long-standing tension via a novel logic, Quantitative Linear Logic (QLL), with foundational ambitions. Our design is driven by naturality -- the idea that, since logical constraints are translated to losses, the semantics of the connectives should be pertinent operations used in ML practice (that is, sum and log-sum-exp) on additive quantities (like logits). We then judge the result on two aspects: logical adequacy -- that they satisfy most of the standard logical laws of Linear Logic; and empirical effectiveness -- test-time performance (as measured by adversarial attacks) is well-correlated to the actual verification of the logical constraints (as measured by off-the-shelf neural network verifiers), which makes QLL stand out among SoTA techniques.