This work presents a collection of multi-modal domain adapted large language models built upon the instruction tuned variants of open weight models (Gemma 3, Qwen 3, Gemma 4) using a relatively small dataset of around 50 million tokens. The dataset consists of open-access additive manufacturing journal articles with data extracted for the domain adaptive pretraining and visual instruction tuning processes. Various stages of the developed model are evaluated with the Additive-Manufacturing-Benchmark which consists of additive manufacturing domain specific tasks compiled published resources. Domain adapted and instruction tuned models exhibit proficiency in both language and vision based tasks, achieving accuracies upwards of 90% in general additive manufacturing knowledge. This domain adaptive pretraining and instruction tuning strategy outline an accessible specialization method for large language models to a domain such as additive manufacturing.

Epistemic reasoning requires agents to infer the state of the world from partial observations and information about other agents' knowledge. Prior work evaluating LLMs on epistemic puzzles often frames failures as memorization rather than reasoning. We argue that this dichotomy is too coarse for newer models: memorization is a limiting case of pattern-based reasoning, where a model matches a task to a familiar template and applies the corresponding solution. We introduce a two-dimensional benchmark over DEL-style puzzles, separating narrative familiarity from inference complexity, allowing us to distinguish pattern-based from epistemic reasoning. We find that models are substantially more robust to surface form changes than prior work suggested, yet consistently struggle in asymmetric settings where familiar patterns no longer apply and success requires tracking fragmented epistemic states.

Speech emotion recognition (SER) systems can exhibit gender-related performance disparities, but how such bias manifests in multilingual speech LLMs across languages and modalities is unclear. We introduce a novel multilingual, multimodal benchmark built on MELD-ST, spanning English, Japanese, and German, to quantify language-specific SER performance and gender gaps. We find bias is strongly language-dependent, and multimodal fusion does not reliably improve fairness. To address these, we propose ERM-MinMaxGAP, a fairness-informed training objective, which augments empirical risk minimization (ERM) with a proposed adaptive fairness weight mechanism and a novel MinMaxGAP regularizer on the maximum male-female loss gap within each language and modality. Building upon the Qwen2-Audio backbone, our ERM-MinMaxGAP approach improves multilingual SER performance by 5.5% and 5.0% while reducing the overall gender bias gap by 0.1% and 1.4% in the unimodal and multimodal settings, respectively.

Understanding hand-object interaction (HOI) is fundamental to computer vision, robotics, and AR/VR. However, conventional hand videos often lack essential physical information such as contact forces and motion signals, and are prone to frequent occlusions. To address the challenges, we present Glove2Hand, a framework that translates multi-modal sensing glove HOI videos into photorealistic bare hands, while faithfully preserving the underlying physical interaction dynamics. We introduce a novel 3D Gaussian hand model that ensures temporal rendering consistency. The rendered hand is seamlessly integrated into the scene using a diffusion-based hand restorer, which effectively handles complex hand-object interactions and non-rigid deformations. Leveraging Glove2Hand, we create HandSense, the first multi-modal HOI dataset featuring glove-to-hand videos with synchronized tactile and IMU signals. We demonstrate that HandSense significantly enhances downstream bare-hand applications, including video-based contact estimation and hand tracking under severe occlusion.

Autoregressive models trained via next-token prediction implicitly learn the conditional independence structure of their data-generating process. We exploit this observation to perform scalable causal discovery from a single observed sequence of discrete events -- without any task-specific retraining. Such single-stream settings arise naturally in vehicle diagnostics, manufacturing systems, and patient trajectories, yet they remain largely unsolved: the absence of repeated samples, massive event vocabularies, and long-range temporal dependencies render existing methods either inaccurate or computationally intractable. We introduce TRACE, a framework that repurposes any pretrained autoregressive model as a density estimator for conditional mutual information, the fundamental primitive for conditional independence testing. By constructing parallelized CI tests on GPUs, TRACE recovers both the sample-level time causal graph and its summary projection, scaling linearly with the vocabulary size while naturally handling delayed causal effects. Crucially, we prove that minimizing the standard cross-entropy pretraining loss directly minimizes an upper bound on the causal identification error, establishing a duality between sequence prediction and causal discovery. On nonlinear SCMs (|X| = 8000) and real-world vehicle diagnostic logs (|X| = 29100), TRACE is the first applicable method at this scale, outperforming the strongest baseline by over 20 F1 points.

Despite rapid progress in building conversational AI agents, robustness is still largely untested. Small shifts in user behavior, such as being more impatient, incoherent, or skeptical, can cause sharp drops in agent performance, revealing how brittle current AI agents are. Today's benchmarks fail to capture this fragility: agents may perform well under standard evaluations but degrade spectacularly in more realistic and varied settings. We address this robustness testing gap by introducing TraitBasis, a lightweight, model-agnostic method for systematically stress testing AI agents. TraitBasis learns directions in activation space corresponding to steerable user traits (e.g., impatience or incoherence), which can be controlled, scaled, composed, and applied at inference time without any fine-tuning or extra data. Using TraitBasis, we extend $τ$-Bench to $τ$-Trait, where user behaviors are altered via controlled trait vectors. We observe on average a 2%-30% performance degradation on $τ$-Trait across frontier models, highlighting the lack of robustness of current AI agents to variations in user behavior. Together, these results highlight both the critical role of robustness testing and the promise of TraitBasis as a simple, data-efficient, and compositional tool. By powering simulation-driven stress tests and training loops, TraitBasis opens the door to building AI agents that remain reliable in the unpredictable dynamics of real-world human interactions. We have open-sourced $τ$-Trai across four domains: airline, retail, telecom, and telehealth, so the community can systematically QA their agents under realistic, behaviorally diverse intents and trait scenarios: https://github.com/collinear-ai/tau-trait.

Pedestrian detection in RGB images is a key task in pedestrian safety, as the most common sensor in autonomous vehicles and advanced driver assistance systems is the RGB camera. Low-light pedestrian detection lacks large public datasets and autolabelling pipelines. This research proposes a solution in the form of an automated infrared-RGB pipeline. The pipeline consists of 1) Infrared detection, where a fine-tuned model for infrared pedestrian detection is used 2) Label transfer process from the infrared detections to their RGB counterparts 3) Training object detection models using the generated labels for low-light RGB pedestrian detection. The research was performed using the KAIST dataset. For evaluation, three object detection models, DETR, YOLO, and RCNN, were trained on generated and ground truth labels. When compared on previously unseen images, the results showed that the models trained on generated labels out-performed the ones trained on ground-truth in 5 out of 6 cases for the mAP@50 and LAMR metrics, and outperformed ground-truth on mAP@50-95 in all cases. Acquired results indicate that the proposed auto-labelling pipeline could be used for scalable annotation of low-light datasets for pedestrian detection. The source code for this research is available on GitHub: https://github.com/BouzoulasDimitrios/IR-RGB-autoamed-low-light-pedestrian-labelling

Three-layer neural networks are known to form singular learning models, and their Bayesian asymptotic behavior is governed by the learning coefficient, or real log canonical threshold. Although this quantity has been clarified for regular models and for some special singular models, broadly applicable methods for evaluating it in neural networks remain limited. Recently, a formula for the local learning coefficient of semiregular models was proposed, yielding an upper bound on the learning coefficient. However, this formula applies only to nonsingular points in the set of realization parameters and cannot be used at singular points. In particular, for three-layer neural networks, the resulting upper bound has been shown to differ substantially from learning coefficient values already known in some cases. In this paper, we derive a formula for an upper bound on local learning coefficients at a class of singular realization parameters in three-layer neural networks. This formula can be interpreted as a counting rule under budget, demand, and supply constraints. In the non-polynomial real-analytic case, the formula applies in general settings, whereas in the polynomial case it applies under the restriction that the true distribution has no hidden units. In particular, our result covers activation functions such as the swish function and also includes polynomial activation functions under the above restriction, thereby extending previous results to a broader class of activation functions. We further show that, when the input dimension is one, the numerical value given by the right-hand side of our upper-bound formula agrees with the previously known learning coefficient, thereby providing a useful comparison with known exact results. Our result also provides a systematic perspective on how the weight parameters of three-layer neural networks affect the learning coefficient.

Scalable embodied intelligence is constrained by the scarcity of diverse, long-horizon robotic manipulation data. Existing video world models in this domain are limited to synthesizing short clips of simple actions and often rely on manually defined trajectories. To this end, we introduce MIND-V, a cognitive hierarchical world model designed to synthesize physically plausible and logically coherent videos of long-horizon robotic manipulation. Inspired by cognitive science, MIND-V bridges high-level reasoning with pixel-level synthesis through three core components: a Semantic Reasoning Hub (SRH) that leverages a pre-trained vision-language model for task planning; a Behavioral Semantic Bridge (BSB) that translates abstract instructions into domain-invariant representations; and a Motor Video Generator (MVG) for conditional video rendering. MIND-V employs Staged Visual Future Rollouts, a test-time optimization strategy to enhance long-horizon robustness. To enforce adherence to physical laws, we introduce a GRPO reinforcement learning post-training phase guided by a novel Physical Foresight Coherence (PFC) reward. PFC leverages the V-JEPA2 world model as a physics referee to penalize implausible dynamics in the latent feature space. Experiments confirm MIND-V's SOTA performance in long-horizon simulation and its significant value for policy learning, introducing a scalable and fully autonomous framework for embodied data synthesis.

Real-time, on-device segmentation is critical for latency-sensitive and privacy-aware applications such as smart glasses and Internet-of-Things devices. We introduce PicoSAM3, a lightweight promptable visual segmentation model optimized for edge and in-sensor execution, including deployment on the Sony IMX500 vision sensor. PicoSAM3 has 1.3M parameters and combines a dense CNN architecture with region of interest prompt encoding, Efficient Channel Attention, and knowledge distillation from SAM2 and SAM3. On COCO and LVIS, PicoSAM3 achieves 65.45% and 64.01% mIoU, respectively, outperforming existing SAM-based and edge-oriented baselines at similar or lower complexity. The INT8 quantized model preserves accuracy with negligible degradation while enabling real-time in-sensor inference at 11.82ms latency on the IMX500, fully complying with its memory and operator constraints. Ablation studies show that distillation from large SAM models yields up to +14.5% mIoU improvement over supervised training and demonstrate that high-quality, spatially flexible promptable segmentation is feasible directly at the sensor level.

Evaluating 'anime-like' voices currently relies on costly subjective judgments, yet no standardized objective metric exists. A key challenge is that anime-likeness, unlike naturalness, lacks a shared absolute scale, making conventional Mean Opinion Score (MOS) protocols unreliable. To address this gap, we propose AnimeScore, a preference-based framework for automatic anime-likeness evaluation via pairwise ranking. We collect 15,000 pairwise judgments from 187 evaluators with free-form descriptions, and acoustic analysis reveals that perceived anime-likeness is driven by controlled resonance shaping, prosodic continuity, and deliberate articulation rather than simple heuristics such as high pitch. We show that handcrafted acoustic features reach a 69.3% AUC ceiling, while SSL-based ranking models achieve up to 90.8% AUC, providing a practical metric that can also serve as a reward signal for preference-based optimization of generative speech models.

Industrial Control Systems (ICS) underpin critical infrastructure and face growing cyber-physical threats due to the convergence of operational technology and networked environments. While machine learning-based anomaly detection approaches in ICS shows strong theoretical performance, deployment is often limited by poor explainability, high false-positive rates, and sensitivity to evolving system behavior, i.e., baseline drifting. We propose a Spatio-Temporal Attention Graph Neural Network (STA-GNN) for unsupervised and explainable anomaly detection in ICS that models both temporal dynamics and relational structure of the system. Sensors, controllers, and network entities are represented as nodes in a dynamically learned graph, enabling the model to capture inter-dependencies across physical processes and communication patterns. Attention mechanisms provide influential relationships, supporting inspection of correlations and potential causal pathways behind detected events. The approach supports multiple data modalities, including SCADA point measurements, network flow features, and payload features, and thus enables unified cyber-physical analysis. To address operational requirements, we incorporate a conformal prediction strategy to control false alarm rates and monitor performance degradation under drifting of the environment. Our findings highlight the possibilities and limitations of model evaluation and common pitfalls in anomaly detection in ICS. Our findings emphasise the importance of explainable, drift-aware evaluation for reliable deployment of learning-based security monitoring systems.

Persian poetry plays an active role in Iranian cultural practice, where verses by canonical poets such as Hafez are frequently quoted, paraphrased, or completed from partial cues. Supporting such interactions requires language models to engage not only with poetic meaning but also with culturally canonical surface form. We introduce GhazalBench, a benchmark for evaluating how large language models (LLMs) interact with Persian ghazals under usage-grounded conditions. Unlike prior work that primarily studies memorization as a liability, GhazalBench examines settings where access to exact surface form is functionally important for culturally grounded interaction. The benchmark evaluates two complementary abilities: poem-to-prose understanding and canonical surface-form access under varying semantic and lexical cues. Across several proprietary and open-weight multilingual LLMs, we observe a consistent dissociation: models generally capture poetic meaning but struggle to produce exact verse completions in open-ended settings, while recognition-based settings substantially reduce this gap. Parallel experiments on English sonnets show markedly stronger completion performance, suggesting that these limitations are tied more to differences in training exposure than to inherent architectural constraints. Our findings highlight the need for evaluation frameworks that jointly assess meaning, form, and cue-dependent access to culturally significant texts. GhazalBench is available at https://anonymous.4open.science/r/GhazalBench/.

Similarities between language representations derived from Self-Supervised Speech Models (S3Ms) have been observed to primarily reflect geographic proximity or surface typological similarities driven by recent expansion or contact, potentially missing deeper genealogical signals. We investigate how scaling an S3M-based language identification system from 126 to 4,017 languages reshapes this topology, and find a non-linear effect: phylogenetic recovery stays flat up to the 1K scale, but the 4K model undergoes a qualitative shift, resolving both clear lineages and long-term linguistic contact. Most strikingly, a robust Pacific macro-cluster emerges, grouping genealogically unrelated Papuan, Oceanic, and Australian languages, and we trace its driver to a concentrated encoding that captures shared acoustic signatures such as global energy dynamics. These results suggest that massive S3Ms internalize multiple layers of language history, offering a promising perspective for computational phylogenetics and the study of language contact.

Hierarchical policies decompose language-conditioned long-horizon robotic manipulation into a high-level planner and a low-level controller. However, effective coordination between HL and LL requires that both components operate on compatible subgoal distributions. We propose ORCHID, a self-training framework that enables stable online improvement of hierarchical diffusion policies by aligning planning and control through iterative refinement. By filtering policy samples via environment feedback, ORCHID identifies trajectories where the planner and controller are jointly successful and distills them back into both modules via supervised learning. This process induces a bidirectional co-adaptation: the planner grounds its subgoals in the actual reaching capabilities of the controller, while the controller specializes in the trajectory structures the planner produces. By relying on supervised distillation of filtered on-policy samples, ORCHID avoids the instability typical of online hierarchical gradient-based RL training with diffusion models. On the CALVIN benchmark, ORCHID allows a lightweight, initially weak model to outperform pure offline methods, including a Vision-Language-Action model twice its size.

Multi-step theorem prediction is a central challenge in geometry problem solving. Existing neural-symbolic approaches rely heavily on supervised parametric models, which exhibit limited generalization to evolving theorem libraries. In this work, we explore training-free theorem prediction through the lens of in-context learning (ICL). We identify a critical scalability bottleneck, termed Structural Drift: as reasoning depth increases, the performance of vanilla ICL degrades sharply, often collapsing to near zero. We attribute this failure to the LLM's inability to recover latent topological dependencies, leading to unstructured exploration. To address this issue, we propose Theorem Precedence Graphs, which encode temporal dependencies from historical solution traces as directed graphs, and impose explicit topological constraints that effectively prune the search space during inference. Coupled with retrieval-augmented graph construction and a stepwise symbolic executor, our approach enables LLMs to act as structured planners without any gradient-based optimization. Experiments on the FormalGeo7k benchmark show that our method achieves 89.29% accuracy, substantially outperforming ICL baselines and matching state-of-the-art supervised models. These results indicate that explicit structural priors offer a promising direction for scaling LLM-based symbolic reasoning.

Existing speech anti-spoofing benchmarks rely on a narrow set of public models, creating a substantial gap from real-world scenarios in which commercial systems employ diverse, often proprietary APIs. To address this issue, we introduce MultiAPI Spoof, a multi-API audio anti-spoofing dataset comprising about 230 hours of synthetic speech generated by 30 distinct APIs, including commercial services, open-source models, and online platforms. Furthermore, we propose Nes2Net-LA, a local-attention enhanced variant of Nes2Net that improves local context modeling and fine-grained spoofing feature extraction. Based on this dataset, we also define the API tracing task, enabling fine-grained attribution of spoofed audio to its generation source. Experiments show that Nes2Net-LA achieves state-of-the-art performance and offers superior robustness, particularly under diverse and unseen spoofing conditions. Code \footnote{https://github.com/XuepingZhang/MultiAPI-Spoof} and dataset \footnote{https://xuepingzhang.github.io/MultiAPI-Spoof-Dataset/} have been released.

In clinical tabular prediction, classical machine learning models with feature engineering often outperform neural methods. LLMs are increasingly used to automate this process, acting as domain experts that propose diverse feature transformations to boost downstream performance. However, existing LLM-based methods decouple feature generation from the downstream model: the LLM receives no signal about which features currently drive predictions or where the model's representational capacity falls short, so proposals are neither targeted to promising regions of the feature space nor tailored to the learner's inductive bias. This shortcoming is amplified in healthcare data, which simultaneously exhibits class imbalance, heterogeneous feature spaces, and strict interpretability requirements. In this paper, we propose MedFeat, the first feature engineering framework inspired by the workflow of machine learning practitioners, leveraging model-awareness and feature importance signals to iteratively guide feature discovery for clinical tabular learning. We evaluate MedFeat on a broad range of challenging real-world clinical tasks and show that it statistically significantly outperforms state-of-the-art baselines, with an average improvement of more than 10% over the baseline across models with distinct inductive biases.

Continuum and soft robots can transform automation tasks requiring compliant interaction in constrained or unstructured environments, including healthcare, agriculture, marine, and space applications. However, their complex mechanics introduce significant challenges in modeling and control. Low-dimensional continuum mechanical models, such as rod theories, effectively capture the large deformations of slender bodies in contact-rich scenarios while balancing accuracy and computational efficiency. This paper presents a vertical survey of rod models for continuum and soft robots, spanning their mathematical foundations, robot modeling, and control applications. We review the main rod theories adopted in soft robotics and introduce a deformation-based classification of rod models for continuum and soft robots. Furthermore, we survey recent model-based and learning-based control strategies leveraging rod models, highlighting their role in manipulation and physical interaction tasks. Finally, we discuss advantages, limitations, research gaps, and emerging directions of rod-based approaches. This paper aims to serve as a reference for developing models and control strategies for continuum and soft robots.

General-purpose simulators have accelerated the development of robots. Traditional simulators based on first-principles, however, typically require full-state observability or depend on parameter search for system identification. This work presents \texttt{CableRobotGraphSim}, a novel Graph Neural Network (GNN) model for cable-driven robots that aims to address shortcomings of prior simulation solutions. By representing cable-driven robots as graphs, with the rigid-bodies as nodes and the cables and contacts as edges, this model can quickly and accurately match the properties of other simulation models and real robots, while ingesting only partially observable inputs. Accompanying the GNN model is a sim-and-real co-training procedure that promotes generalization and robustness to noisy real data. This model is further integrated with a Model Predictive Path Integral (MPPI) controller for closed-loop navigation, which showcases the model's speed and accuracy.

Kuzushiji was one of the most widely used cursive writing systems in pre-modern Japan. Due to its highly cursive forms and extensive glyph variations, most modern Japanese readers are unable to read Kuzushiji characters. Consequently, recent studies have focused on developing automated Kuzushiji character recognition (KCR) methods, which have achieved strong performance on relatively clean Japanese historical document images. Although seals frequently appear in Japanese historical documents, existing methods often fail to maintain recognition accuracy under seal interference, particularly when seals overlap with characters. To address this challenge, we propose a seal-robust KCR framework. Based on character detection, classification, and ordering, the proposed framework additionally incorporates document restoration to mitigate seal interference, thereby improving overall recognition performance. In addition, we introduce a novel synthetic data augmentation strategy to enhance the performance of character detection models. We further correct annotation errors, reconstruct the dataset, and create a synthetic test set to simulate severe seal interference. Experimental results demonstrate the effectiveness of the proposed framework in mitigating the impact of seal interference on KCR. Compared with a conventional baseline and NDLkotenOCR, it achieves relative character error rate (CER) reductions of 39.7% and 5.9%, respectively, on the real test set, and 50.1% and 41.7%, respectively, on the synthetic test set.

Loss spikes remain a persistent obstacle in large-scale language model pretraining. While previous research has attempted to identify the root cause of loss spikes by investigating individual factors, we observe that, in practice, such spikes are typically triggered by the confluence of heterogeneous factors. Empirically, loss spikes may arise from a combination of data outliers, hardware or transient computational faults, numerical precision issues, and hyperparameter settings. Regardless of the underlying cause, these spikes manifest as unstable optimizer updates, as abnormal gradients contaminate both first- and second-moment states. In this paper, we propose a principled gradient-centric remedy: AdaGC, an adaptive per-tensor gradient clipping scheme that mitigates such contamination by bounding gradient norms relative to a tensor-wise exponential moving average of their historical clipped values. AdaGC is optimizer-agnostic, introduces negligible memory overhead, and reduces communication costs compared to GlobalGC, particularly in hybrid-parallel distributed training. Experiments on Llama-2 7B, Mixtral 8x1B, and ERNIE 10B-A1.4B demonstrate that AdaGC robustly eliminates training instabilities, consistently reducing spike scores to zero for all models and improving downstream accuracy over GlobalGC by 1.32%, 1.27%, and 2.48%, respectively. Furthermore, AdaGC seamlessly integrates with optimizers such as Muon and Lion, consistently yielding higher average accuracy and zero spike scores. The code is available at https://github.com/PaddlePaddle/PaddleFleet (see Research/AdaGC).

Traditional neural topic models are typically optimized by reconstructing the document's Bag-of-Words (BoW) representations, overlooking contextual information and struggling with data sparsity. In this work, we introduce a novel topic model training framework by Distilling Soft Labels (DSL) from Language Models (LMs). To construct the contextually enriched reconstruction signals, we project the next token probabilities, conditioned on a specialized prompt, onto a pre-defined vocabulary, and train the topic models to reconstruct the soft labels using the LM hidden states. This produces higher-quality topics that are more closely aligned with the underlying thematic structure of the corpus. Extensive experiments demonstrate that DSL achieves substantial improvements in topic coherence and assignment accuracy over existing baselines. Additionally, we also introduce a retrieval-based metric, which shows that our approach significantly outperforms existing methods in identifying semantically similar documents, highlighting its effectiveness for retrieval-oriented applications.

Time series anomaly detection is critical in many real-world applications, where effective solutions must localize anomalous regions and support reliable decision-making under complex settings. However, most existing methods frame anomaly detection as a purely discriminative prediction task with fixed feature representations, rather than an evidence-driven diagnostic process. As a result, they often struggle when anomalies exhibit strong context dependence, diverse patterns, or domain shifts across datasets. To address these challenges, we propose AnomaMind, an agentic time series anomaly detection framework that reformulates anomaly detection as a sequential decision-making process. AnomaMind operates through a coarse-to-fine workflow that first localizes suspicious intervals, then constructs diagnostic evidence through tool interaction, and finally refines anomaly decisions through self-reflection. The workflow is supported by a toolkit box that combines knowledge memory and numerical diagnostics: visual anomaly patterns mined from training data and domain knowledge provide contextual guidance, while statistical, value-based, change-based, and region-level operators provide measurable evidence for verification. AnomaMind further adopts a hybrid inference mechanism in which general-purpose models handle flexible reasoning, tool invocation, and refinement, while a detection-specific policy is optimized with rule-based rewards for parsable outputs, F1-score alignment, and false-positive control. Extensive experiments under both in-domain and cross-domain settings demonstrate that AnomaMind consistently improves anomaly detection performance and enhances generalization across heterogeneous anomaly patterns, validating the effectiveness of tool-augmented reasoning for anomaly detection. The code is available at https://github.com/Xiaoyu-Tao/AnomaMind-TS.

Understanding how and why large language models (LLMs) fail is becoming a central challenge as models rapidly evolve and static evaluations fall behind. While automated probing has been enabled by dynamic test generation, existing approaches often discover isolated failure cases, lack principled control over exploration, and provide limited insight into the underlying structure of model weaknesses. We propose ProbeLLM, a benchmark-agnostic automated probing framework that elevates weakness discovery from individual failures to structured failure modes. ProbeLLM formulates probing as a hierarchical Monte Carlo Tree Search, explicitly allocating limited probing budgets between global exploration of new failure regions and local refinement of recurring error patterns. By restricting probing to verifiable test cases and leveraging tool-augmented generation and verification, ProbeLLM grounds failure discovery in reliable evidence. Discovered failures are further consolidated into interpretable failure modes via failure-aware embeddings and boundary-aware induction. Across diverse benchmarks and LLMs, ProbeLLM reveals substantially broader, cleaner, and more fine-grained failure landscapes than static benchmarks and prior automated methods, supporting a shift from case-centric evaluation toward principled weakness discovery.

Deep learning models for clinical event prediction on electronic health records (EHR) often suffer performance degradation when deployed under different data distributions. While domain adaptation (DA) methods can mitigate such shifts, their "black-box" nature prevents widespread adoption in clinical practice where transparency is essential for trust and safety. We propose ExtraCare to decompose patient representations into invariant and covariant components. By supervising these two components and enforcing their orthogonality during training, our model preserves label information while exposing domain-specific variation at the same time for more accurate predictions than most feature alignment models. More importantly, it offers human-understandable explanations by mapping sparse latent dimensions to medical concepts and quantifying their contributions via targeted ablations. ExtraCare is evaluated on two real-world EHR datasets across multiple domain partition settings, demonstrating superior performance along with enhanced transparency, as evidenced by its accurate predictions and explanations from extensive case studies.

Benchmarks establish a standardized evaluation framework to systematically assess the performance of large language models (LLMs), facilitating objective comparisons and driving advancements in the field. However, existing benchmarks fail to differentiate question difficulty, limiting their ability to effectively distinguish models' capabilities. To address this limitation, we propose RankLLM, a novel framework designed to quantify both question difficulty and model competency. RankLLM introduces difficulty as the primary criterion for differentiation, enabling a more fine-grained evaluation of LLM capabilities. RankLLM's core mechanism facilitates bidirectional score propagation between models and questions. The core intuition of RankLLM is that a model earns a competency score when it correctly answers a question, while a question's difficulty score increases when it challenges a model. Using this framework, we evaluate 30 models on 35,550 questions across multiple domains. RankLLM achieves 90% agreement with human judgments and consistently outperforms strong baselines such as IRT. It also exhibits strong stability, fast convergence, and high computational efficiency, making it a practical solution for large-scale, difficulty-aware LLM evaluation.

Generative sequence modeling faces a fundamental tension between the expressivity of Transformers and the efficiency of linear sequence models. Existing efficient architectures are theoretically bounded by shallow, single-step linear updates, while powerful iterative methods like Test-Time Training (TTT) break hardware parallelism due to two dimensions of serial dependency: token-level state reliance and step-level iteration loops. We propose PRISM (Parallel Residual Iterative Sequence Model) to resolve this tension. PRISM explicitly reconstructs the expressive gate x residual x direction iteration pattern of TTT in a parallelizable form. We employ a Write-Forget Decoupling strategy that isolates non-linearity within the injection operator. To bypass the serial dependency of explicit solvers, PRISM utilizes a two-stage proxy architecture: a short-convolution anchors the initial residual using local history energy, while a learned predictor estimates the refinement updates directly from the input. This design distills structural patterns associated with iterative correction into a parallelizable feedforward operator. Theoretically, we prove that this formulation achieves Rank-$L$ accumulation, structurally expanding the update manifold beyond the single-step Rank-$1$ bottleneck. Empirically, it achieves comparable performance to explicit optimization methods while achieving \textbf{174x higher throughput}. Codes are available in https://github.com/gpr-prism/prism/.

Contemporary visuo-motor dexterity models often rely on expressive policy classes with diffusion and transformer backbones to achieve strong performance. However, these architectures require significant data and computational resources, and remain far from reliable, particularly for multi-fingered dexterity. Importantly, they model skills as reactive mappings and rely on fixed-horizon action chunking, creating a rigid trade-off between temporal coherence and reactivity. To address these issues, we first introduce Unified Behavioral Models (UBMs), a framework to represent dexterous skills as coupled dynamical systems that capture how visual features of the environment (visual flow) and proprioceptive states of the robot (action flow) co-evolve. As such, UBMs ensure temporal coherence by construction rather than heuristic averaging. Unlike world models that attempt to predict the impact of arbitrary robot actions on the environment, UBMs target behavioral dynamics that encode how demonstrated robot behavior is related to desired impacts on the environment. A UBM can be viewed as a pseudo planner: given an initial condition, it computes the desired robot behavior over the entire skill horizon, while simultaneously ``imagining" the resulting flow of visual features. To operationalize UBMs, we propose Koopman-UBM, a first instantiation of UBMs as a structured latent linear system. K-UBM is computationally efficient, enabling reactivity and adaptation via an online replanning strategy: the model acts as its own runtime monitor, automatically triggering replanning when predicted and observed visual flow diverge beyond a threshold. Across seven simulated tasks and four real-world tasks, our approach matches or exceeds the performance of state-of-the-art baselines, while offering considerably faster inference, smooth execution, robustness to occlusions, and flexible replanning.

Scientific diagrams convey explicit structural information, yet modern text-to-image models often produce visually plausible but structurally incorrect results. Existing benchmarks either rely on image-centric or subjective metrics insensitive to structure, or evaluate intermediate symbolic representations rather than final rendered images, leaving pixel-based diagram generation underexplored. We introduce SciFlow-Bench, a structure-first benchmark for evaluating scientific diagram generation directly from pixel-level outputs. Built from real scientific PDFs, SciFlow-Bench pairs each source framework figure with a canonical ground-truth graph and evaluates models as black-box image generators under a closed-loop, round-trip protocol that inverse-parses generated diagram images back into structured graphs for comparison. This design enforces evaluation by structural recoverability rather than visual similarity alone, and is enabled by a hierarchical multi-agent system that coordinates planning, perception, and structural reasoning. Experiments show that preserving structural correctness remains a fundamental challenge, particularly for diagrams with complex topology, underscoring the need for structure-aware evaluation.