Papers

Post-training INT8 quantization is typically validated by checking accuracy retention, which requires ground-truth labels often unavailable in low-resource clinical settings. We show a label-free alternative: comparing a model's own output before and after quantization. We compute cosine similarity between the FP32 and INT8 models' output probability vectors for each prediction, which requires only a single forward pass. This output-agreement similarity separates lower-accuracy predictions across four evaluated architectures (ResNet-50, VGG-16-BN, Wide ResNet-50, and ViT-B/16) under fixed-percentile splits, outperforming a confidence-only baseline for all four. These results show that quantized models can be compared to their baselines to assess their drift and determine whether they are well suited for clinical deployment.

The stories told about artificial intelligence (AI) play a central role in shaping public understanding, policy debates, and expectations about technological change. News media are especially influential in this process. Using a cross-regional corpus of 12,924 English-language news articles about AI published since 2022, we examine how storytelling about AI varies across media systems in four regions: North America, China, India, and the Middle East. Leveraging large language models, we extract the implicit narrative lessons conveyed by each article—commonly known as story morals—capturing value-laden interpretations that extend beyond thematic content or sentiment. We find that North American AI coverage is uniquely characterized by narratives of caution, while under-emphasizing narratives of innovation and collaboration that are more prevalent elsewhere. However, this pattern is largely explained by a broader regional negativity bias in North American news sentiment, rather than by heightened concern about AI itself. Our findings bring into view the culturally conditioned nature of AI news narratives.

Automated post-training transforms pre-trained large language models (LLMs) into models capable of handling diverse domains by automatically constructing and optimizing a recipe of coupled decisions over data sources, data processing, cross-domain composition, and training configuration under a limited experimental budget. Existing automation either searches predefined recipe spaces that enable systematic optimization but remain fixed rather than adapting their available choices to the target tasks, or uses open-ended agents that discover task-specific choices but leave them implicit in trial-level edits, limiting their systematic reuse and recombination across trials. To fill this gap, we introduce optimization over a task-adapted recipe space, a novel paradigm that can adapt the recipe space itself to the target tasks while representing its choices explicitly for systematic reuse, recombination, and joint optimization. We instantiate this paradigm with Structured Recipe Optimization (SRO), comprising an LLM-based Recipe-Space Designer that constructs the task-adapted recipe space and a Multi-Domain Recipe Optimizer that uses structured trial evidence to guide LLM candidate construction and multi-objective Bayesian trial selection. Extensive experiments show that SRO achieves the highest average score in all evaluated target-model and agent-LLM combinations. In joint post-training across three capability domains and six benchmarks, SRO reaches 41.1 versus 15.4 for the strongest baseline on Qwen3-1.7B with GPT-5.4, and 53.8 versus 40.8 on Qwen3-4B with DeepSeek-V4-Flash. Our code is available at https://github.com/RuohLiuq/SRO.

Different MRI modalities reveal different parts of a brain tumor. Segmentation models combine information from these modalities to outline tumor regions. Many models use a separate encoder for each modality. However, exchanging features among these encoders at several stages can increase model complexity. We introduce the Block-Wise Shared Adapter (BWSA), which combines compressed features from the modality encoders in a shared path and returns adapted features to each stream. In its fixed-blending form, BWSA adds the adapted features to the original features. Its trainable-gated variant (TG-BWSA) instead learns separate weights for the original and adapted features at each modality and encoder stage. We evaluate both variants on BraTS 2020 and BraTS 2023 with four segmentation backbones. On the official BraTS 2023 online validation set, BWSA increases average region-wise Dice by 1.3–7.6 percentage points relative to the corresponding unmodified backbones. TG-BWSA increases it by 2.8–7.7 points. Both variants add approximately 0.85–3.73% parameters and reduce average HD95 relative to the unmodified backbones.

Latent planners often rank candidate action sequences by terminal Euclidean distance to a goal latent. Exact agreement with transition distance would make this score ideal, but is stronger than planning requires: along a transition-shortest rollout, the essential requirement is that latent distance does not reverse true transition progress. We show that this pathwise no-reversal requirement is precisely the classical self-approaching condition. To the best of our knowledge, we are the first to identify self-approachingness as a planner-facing compatibility condition for terminal Euclidean latent planning and to connect its violation to an exact finite-horizon candidate-order reversal. Because visual world models predict discrete latent sequences, we analyze their polygonal interpolation. We prove that an acute bend creates a nonempty region in which terminal Euclidean distance ranks a transition-worse candidate ahead of one that makes genuine progress, even under exact rollouts. We further interpret latent dimension as angular capacity through a classical spherical-code bound. A controlled six-branch experiment, including a zero-padded three-dimensional baseline, validates the predicted ordering failure and separates angular arrangement from ambient dimensionality.

Vision language models (VLMs) are used in creative tasks such as visual storytelling, design ideation, art interpretation, and advertising. Yet we lack principled ways to measure the lateral reasoning of these models compared to human potential. We introduce CREATE, a semi-automatically constructed benchmark containing 155 visual Remote Associates Test problems. These problems require a solver to find an underlying connection among three seemingly dissimilar images. We conduct a human study (N=25) to validate and ground the quality of our benchmark, and evaluate 11 VLMs. Our findings show that most VLMs underperform humans on this task. Our experiments indicate that the associative ability of VLMs depends on the cue concepts provided in the text and does not universally extend to visual puzzles.

Vision-language-action (VLA) policies excel at generalist robotic manipulation but generate actions without explicitly modeling their environmental consequences. World action models (WAMs) incorporate future prediction, yet most rely on costly video generation, while latent alternatives either use predicted futures only as auxiliary targets or infer latent actions from suboptimal visual transitions through costly large-scale video pretraining. We introduce WAVE-0 (World-model-Augmented VLA Execution), a framework that bridges VLAs and WAMs through a lightweight latent world model. By connecting the semantic understanding of pretrained VLAs with explicit future prediction, WAVE-0 allows anticipated action consequences to directly guide action generation. Built on a pretrained VLA, its 257M world model translates VLM-generated latent action plans into predicted future representations that condition the action expert, unifying latent planning, future prediction, and action generation. The world model is trained directly on ground-truth robot actions with a counterfactual ranking objective, requiring neither pixel reconstruction nor large-scale video pretraining. WAVE-0 achieves success rates of 91.47% on RoboTwin 2.0 and 98.7% on LIBERO, along with 88.0%, 61.4%, and 75.0% on LIBERO-Plus, LIBERO-Pro, and LIBERO-Para, respectively. On real-world tasks, it improves the success rate of $\pi_{0.5}$ from 38.3% to 73.3%. These results demonstrate strong manipulation performance and generalization across diverse distribution shifts.

Retrieval-Augmented Generation (RAG) has emerged as a dominant paradigm for improving Large Language Model (LLM) by intertwining it with other components, such as a vector/graph database. Therefore, the resulting RAG systems are often highly configurable, under which a non-trivial amount of failures are related to the configurations. However, there remains a lack of systematic understanding regarding the configuration issues for RAG systems. To bridge this gap, in this paper, we conduct the first large-scale empirical study of 654 real-world configuration issues collected from 10 diverse RAG systems. Drawing on this, we provide a complete categorization of the symptoms, root causes, relevant stages, and fix patterns of configuration issues for RAG systems, as well as several new observations. Our findings not only reveal the characteristics of configuration issues, but also strongly link them to the unique pipeline stages of RAG, e.g., we found that most symptoms can be observed at the early or late stage, while the stage of the root cause might not be at the same stage as where the symptom is identified. We also provide nine actionable insights to the practitioners on testing, localizing, and fixing the configuration issues for RAG systems.

Large language model (LLM) routers reduce serving cost by choosing which model to call and how many output tokens to grant it for each query. Existing routers rank model-budget actions using a single quality metric. This does not work for mixed workloads. For ordinary queries, a good answer is correct and useful (utility); for security-sensitive queries, a good answer refuses harmful requests or resists malicious instructions (security). These two objectives prefer different models and budgets. We propose SCRouter, a router that learns when to prioritize security and how to allocate the token budget. For each model-budget pair, it estimates the probability that the response finishes within budget. It uses this probability to predict utility and security separately. A small probe determines how to weight these two predictions for each query. SCRouter then selects the model and budget by balancing the weighted utility and security against cost. We also build SCRouterBench, a benchmark of 38,869 utility and security queries spanning nine security dimensions, each answered under all 144 combinations of 12 models and 12 budgets, plus 2,753 out-of-domain queries. SCRouter improves the baselines by 13-29 points in-domain and 19-23 points out-of-domain.

Simulation of HPC applications is an essential tool for analyzing system design trade-offs and enabling effective hardware–software co-design. As modern supercomputers increasingly rely on GPU acceleration, MPI implementations have evolved to support GPU-aware communication, enabling direct GPU data transfers. Modeling these mechanisms is therefore critical for simulating HPC applications on current systems. In this paper, we introduce Orrery, an interposition layer built on SimGrid's SMPI to model GPU-aware MPI communication using existing collective implementations and flow-based simulation. Orrery intercepts GPU memory allocations, kernel launches, data transfers, and MPI operations, and models their execution in SimGrid according to target system specifications, enabling simulation without requiring physical GPUs. We validate Orrery using CUDA-aware MPI microbenchmarks and a Jacobi solver on up to 32 nodes of the MUSICA supercomputer and demonstrate that it reproduces key GPU-aware MPI mechanisms.

Aerial map matching provides position observations for navigation, but a position error alone does not determine how an observation should be weighted. We define five benchmark tasks: position availability and agreement, grouped error structure, predictive uncertainty, controlled updates, and temporal accumulation. A case study in digital-surface-model-assisted localization evaluates base-anchored uncertainty calibration within this framework. The method combines propagated image and map sensitivities with the second moment of a compact configuration subset relative to the reported position. The reported position remains unchanged, and scale is fitted using separate position references without local surface-height ground truth. Thirteen covariance models are evaluated on two held-out query sets from locations represented in the calibration data. Relative to a constant full-matrix baseline, the method reduces equal-location normalized position-update risk by 7.70% on 728 conditions and 5.46% on a separate 142 conditions under an independent prior, with all localization failures included. On the primary set, equal-location negative log likelihood improves by 2.571 nat. These results show improved reference-based weighting at seen locations. Exploratory base-quality and conditional-addition controls leave the incremental update benefit of configuration disagreement unresolved. Navigation benefit and real temporal accumulation remain untested.

Transient stability constraints alter the economic dispatch of power systems, while the resulting price signals and their implications for market settlement remain insufficiently characterized. Existing approaches identify stability-related components in locational marginal prices, without deriving shadow prices for individual assets, such as generators and loads, or analyzing the resulting settlement. To address this gap, this paper derives shadow prices from a transient stability-constrained optimal power flow in which the differential-algebraic equations governing the system dynamics are embedded through direct discretization. Each shadow price decomposes into a nodal energy component and a stability component, the latter formed by the dual variables of the dynamic equations in which the power of the asset appears. Shadow prices are therefore asset-specific: co-located generators and loads are assigned the same nodal energy component, but can be subject to different stability components when their transient behaviors differ. Case studies on the WECC 9-bus system and a modified IEEE 39-bus system show that settlement at the proposed shadow prices allocates the cost of the stability constraints among the assets without any side payment. An asset that reduces the stability margin is charged its stability component, whereas an asset that improves it is paid, and no generator is dispatched at a loss.

We prove that the last iterate of Optimistic Follow-The-Regularized-Leader (OFTRL) converges at a linear rate to a Nash equilibrium in multiplayer harmonic games. Our result holds under a mild game-regularity condition and for a large family of regularizers, including entropy, Tsallis, and log-barrier. Our analysis is based on a dual perspective of OFTRL: we prove that the interaction between optimism and the global structure of harmonic games leads to the dissipation of an energy function along the dual iterates, for which we establish tight quantitative bounds. Our result also holds for the Extragradient FTRL variant.

Region-of-interest (ROI) selective encryption in H.265/HEVC aims to protect sensitive content while preserving the utility of non-ROI regions. Existing coding-unit (CU)-level approaches still suffer from substantial encryption redundancy for two reasons. First, existing approaches map object-detection bounding boxes to CUs, producing regular rectangular regions, which poorly match the shape of irregular ROI boundaries, such as facial contours. Second, their coding-tree-unit (CTU)-based diffusion isolation expands the protected footprint in fixed $64\times64$ blocks, causing encryption-induced distortion to spread over a large number of non-ROI pixels. To address the aforementioned shortcomings, we propose a prompt segmentation-guided CU-level ROI encryption algorithm, which mainly consists of two core components. The first component is a prompt segmentation-guided fine-grained ROI encryption pipeline. It maps pixel-level ROI masks to CUs according to an ROI-overlap threshold, refines boundary CUs to $8\times8$ granularity, and applies format-compliant encryption to multiple syntax elements in the mapped ROI CUs. The second is a dual diffusion isolation strategy to address the encryption diffusion problem. For spatial isolation, the encoder refines non-ROI boundary regions into $8\times8$ CUs and uses PCM mode to bypass intra-frame prediction. For temporal isolation, it removes motion-vector candidates whose reference blocks overlap ROI CUs. Experimental results demonstrate that the proposed method generates encryption regions that more closely match irregular ROI boundaries, effectively confines distortion to the ROI.

Large Language Models (LLMs) execute billions of floating-point operations per output token across many hardware accelerators. At this scale, hardware-induced faults can bypass system-level detection mechanisms and cause Silent Data Corruption (SDC). Algorithm-Based Fault Tolerance (ABFT) is the state-of-the-art technique for protecting matrix multiplication from such faults, but it incurs non-negligible computational overhead and requires thresholds to distinguish hardware-induced errors from floating-point roundoff. In this paper, we revisit Cauchy–Schwarz Norm Bounds (CSNB) as a low-overhead alternative for detecting SDC during LLM inference. We compare CSNB against ABFT through controlled fault injection and evaluate the runtime overhead of both verification and threshold determination across multiple model sizes, batch sizes, and generation lengths. Our results show that, on average, CSNB retains 99% of the fault detection capability of ABFT for the evaluated dense linear layers while reducing verification overhead by 87% and threshold determination overhead by 67%. For Mixture-of-Experts (MoE) router gate computations, however, CSNB retains only 55% of the fault detection capability of ABFT. These findings suggest that CSNB is a promising low-overhead alternative for protecting dense linear layers in LLM inference, while checksum-based verification remains preferable for the evaluated MoE router gate.

Deep learning for medical image restoration is often constrained by the need for paired training data. Acquiring perfectly registered clean-noisy radiographs in vivo is impractical due to patient motion and radiation-dose limits. To address this, we employ a self-supervised Noise2Noise paradigm that trains on independently corrupted observations without paired clean images. We propose Denoise X, a lightweight self-supervised U-Net model (7.79M parameters) for suppressing quantum mottle and electronic readout noise in low-dose chest radiographs. It is trained on the NIH ChestX-ray14 dataset (112,120 images) using a stochastic Poisson-Gaussian noise simulator. Zero-shot evaluation on the unseen Montgomery County dataset achieved a PSNR of 36.09 dB and SSIM of 0.8806, outperforming the Bilateral filter while preserving high-frequency anatomical details. Using patch-wise inference, Denoise X processes native-resolution chest radiographs in ~2.2 s per image on a Tesla P100 GPU. Cross-domain evaluation on the Shenzhen dataset achieved 36.44 dB PSNR and 0.9099 SSIM, showing strong generalization. For resource-constrained deployment, the model was converted into a 31.16 MB ONNX model for efficient CPU inference on non-DICOM images.

In this paper, we study the last-iterate convergence of clipped stochastic gradient descent under heavy-tailed noise. We assume that the stochastic gradients are unbiased and that the noise has a bounded central $p$-th moment, where $1<p\le2$. For smooth convex objectives, a linearly decreasing learning rate gives a high-probability error bound of order $O(LR^2\log(1/\delta)/T+\sigma R(\log(1/\delta)/T)^{(p-1)/p})$. We further establish last-iterate guarantees for nonsmooth composite objectives and smooth composite objectives with proximal updates. In addition, when a causal adversary replaces at most $q$ gradients, the same clipped update achieves a bound with $\log(1/\delta)$ replaced by $\log(1/\delta)+q$. The corresponding lower bounds for the clean and corrupted oracle models are also provided.

Teacher-free one-step generative models such as W-Flow learn directly from distributional objectives, with the Sinkhorn divergence providing the strongest empirical performance among the energies studied in W-Flow. However, Sinkhorn divergence is based on balanced optimal transport, whose exact marginal constraints can induce undesirable transport under finite minibatch fluctuations. We introduce UW-Flow, which replaces the balanced Sinkhorn energy with a debiased unbalanced Sinkhorn divergence and derives the corresponding Wasserstein velocity. The resulting dynamics retain W-Flow's cross-minus-self structure, recover the balanced limit, and preserve target identifiability under suitable assumptions. We further characterize the finite-batch velocity error of balanced and unbalanced transport by comparing each minibatch field with its population counterpart. Our analysis identifies conditions under which marginal relaxation yields smaller minibatch-to-population velocity error, and shows that for separated matched modes, balanced sensitivity grows with inter-mode separation while the corresponding Unbalanced Optimal Transport (UOT) sensitivity remains controlled by local transport geometry. Controlled experiments on fixed-support and continuous two-mode distributions support these predictions. On MNIST, CIFAR-10, and ImageNet-100, UW-Flow improves generation under matched training budgets, with the largest gains in unconditional generation and consistent improvements in the class-conditional setting.

CMinHash promises a new direction for designing MinHash algorithms with reduced variance and storage space via circulant permutations. However, the original variance analysis is difficult to extend to other MinHash estimators. We present a new approach that gives explicit and easily comparable approximate variance expressions, and use it to analyze the Minner estimator. Numerical experiments show that our variance expression matches the empirical mean square error (MSE) of estimates. Our work thereby demonstrates how to apply circulant permutation to other MinHash estimators and compute their approximate variance.

Generative Artificial Intelligence (GenAI) tools are increasingly integrated into everyday life, including sensitive domains such as religion. Prior works examined alignment with religious values and discussed the potential risks. However, to gain empirical insights into the adoption of these tools in users' religious lives, in this study, we conducted an online survey with 178 Muslim participants in Bangladesh and collected 726 prompts on religious use cases, along with open-ended responses about using these tools in religious contexts. Using reflexive thematic analysis, we identify three key roles these tools play in users' religious lives: (1) religious assistants, (2) perceived religious scholars, and (3) faith counselors. However, participants expressed concerns about misguidance, overreliance, and violations of Islamic tradition, and reported adopting protective strategies to avoid these pitfalls. Based on these, we propose a set of design guidelines for faith-sensitive GenAI systems that would preserve religious boundaries while supporting religious practices.

Language and vision language models can support autonomous systems, but changing operating conditions may compromise the reliability of their semantic outputs. This work introduces the Attribute Reliability Monitor (ARM), a framework that accumulates evidence of suspicion over time to assess individual semantic attributes. ARM accommodates different detectors and temporal monitoring methods, while leaving intervention policies to the consuming system. An initial implementation was evaluated using a specialized vision language model developed for satellite thermal infrared imagery. Monitoring followed a black box approach, using only observable responses without access to internal model states, confidence scores, or external reference answers. Under controlled input degradation, disagreement between responses provided evidence for selective suspension of attributes. The results illustrate how external behavioral monitoring can inform continued use of model outputs, while distinguishing operational suspicion from semantic correctness.

People often tackle difficult problems by thinking more deeply or exploring more possibilities. Similarly, recursive reasoners, which reason by repeatedly applying a learned update rule to latent states, can also scale inference through more updates (depth) or additional trajectories generated by perturbing latent states (width). When people hit a dead end in their thinking, they may step back and approach the problem from a fresh angle. However, neither depth scaling nor perturbation-based width scaling of recursive reasoners can reliably achieve this: both may converge to incorrect attractors. To address this, we propose to explore diverse reasoning traces by applying different update maps to the same latent state, without any additional training. We realize this idea through Input-Conditioned Search (ICS), which searches over input conditions to induce alternative update maps from the pretrained reasoner. Across six reasoning tasks, ICS achieves the highest mean accuracy, while requiring up to two orders of magnitude fewer model FLOPs than the evaluated scaling baselines. Controlled experiments confirm that ICS can redirect trajectories from the same state. Our theoretical analysis and rigorous numerical verification also confirm that ICS can locally eliminate incorrect attracting fixed points. Together, these findings establish ICS as an effective axis for efficient test-time scaling and motivate further research on shaping recursive reasoning during training and at test time.

Foundation models are increasingly used as feature extractors for downstream tasks with limited data, typically through representations obtained from their encoders. However, foundation models trained for structured tasks such as segmentation also form object-specific representations during task execution, whose transferability remains largely unexplored. We investigate Segment Anything Model 2 (SAM2) object pointers, compact object-level representations produced through the segmentation pathway, as transferable representations for downstream tumour prediction. Through experiments on two PET/CT tumour prediction tasks, we find that object pointers learned using segmentation supervision alone already encode tumour-discriminative information, despite never being optimised for the downstream labels. We further introduce task-guided object pointers, where an auxiliary objective jointly shapes the object pointer space for the downstream prediction together with segmentation. Task-guided pointers improve cancer-type classification, outperforming conventional encoder-based transfer and radiomics, while segmentation-only pointers perform best on the more challenging recurrence prediction task. Representation analyses reveal associations between object pointers and interpretable radiomics features, class-dependent organisation of the learned representation space, and explicit spatial grounding around the segmented tumour. Together, these findings suggest that useful transferable representations can emerge through the execution of a foundation model's primary task, providing an alternative to conventional encoder-based transfer.

Language models are increasingly expected to support users in stressful or uncertain situations, where warmth and empathy are desirable. While recent work associates warmer models with reduced reliability, we show that this conclusion is incomplete. Prior studies often train primarily on empathetic dialogue, making it difficult to separate tone effects from SFT-induced distribution shift. We revisit this issue at both the model and evaluation levels. We study prompting, SFT, and RL for inducing warm behavior, and comprehensively evaluate capability preservation using an 11-dataset benchmark suite and EMOFACT under emotional contexts. Across multiple model families, we find that warmth effects are task- and model-dependent, while task-aligned adaptive tone can improve empathy and targeted reliability, although some capability trade-offs remain. We further introduce human-validated empathy evaluators specifically developed to assess empathetic model responses. Our results suggest that models can be more supportive in social interactions while remaining reliable.

The Boolean Satisfiability (SAT) problem is a remarkably expressive computational problem, serving both as a cornerstone of complexity theory and as a unifying language for a wide range of search and reasoning tasks. While $\mathrm{NP}$-complete in general, SAT is well known to be linear-time solvable on instances with bounded treewidth. This motivates the search for "hidden" tree-like structures in formulas—for instance by detecting so-called weak $t$-treewidth backdoors. Such backdoors are sets of $k$ variables that can be assigned in such a way that the remaining instance has treewidth at most $t$ and is satisfiable whenever the original instance is. Ever since Gaspers and Szeider [ICALP 2012] showed that finding a weak $t$-treewidth backdoor for a SAT formula is $\mathrm{W}[2]$-hard with respect to $k$, tractability results have been restricted to structurally sparse formulas—most notably, a linear-time algorithm for $d$-SAT (bounded clause-length SAT) by Fomin, Lokshtanov, Misra, Ramanujan, and Saurabh [SODA 2015] and a non-uniform polynomial-time algorithm when the incidence graph of the input formula belongs to a nowhere-dense graph class (or more generally is $K_{d,d}$-free) by Lokshtanov, Panolan, and Ramanujan [ICALP 2022]. In this paper, we push the tractability boundaries of finding weak $t$-treewidth backdoors beyond sparsity and into the regime of "well-behaved" dense formulas. Aiming towards relaxing structural input restrictions from nowhere-dense to edge-stable graph classes, as featured in the theory of beyond sparsity for FO model checking, or to bounded VC-dimension set systems, we identify signed semi-ladder-freeness as the correct dense structural restriction that establishes the boundary between tractability and intractability along this line.

Understanding agent heterogeneity can inform the design of shared and agent-specific policy representations, however existing criteria provide no unified structural test. We study finite-horizon POSGs and define a type-compatible transformation as a policy-evaluation symmetry when complete policy evaluation commutes with it for every decentralized behavioral policy. Policy-evaluation symmetries form a group for which agent orbits define structural agent types. This classification does not depend on the particular policy chosen or the distribution of initial states. We prove that every componentwise environment automorphism is a policy-evaluation symmetry, but that the converse does not hold. Applying the framework to cooperative Dec-POMDPs shows that agents can belong to different structural types even when they share a common reward. For prescribed weighted reward aggregation, we derive an exact value identity and sufficient conditions for the resulting Dec-POMDP to inherit the policy-evaluation symmetries of the source POSG. Finally, we give sufficient conditions under which repeated model factorizations induce policy-evaluation symmetry, and show that preserving an optimal Boltzmann policy set neither implies nor follows from this symmetry in general.

In the Hunters and Rabbit game, hunters shoot at a set of vertices at each round, while an invisible rabbit must move along an edge. We prove that deciding whether a given number of hunters can guarantee capture is PSPACE-complete, even on simple bipartite graphs. The reduction uses intersection nonemptiness for inverse finite automata: one common word must put all automata in accepting states simultaneously. A clique acting as a safe haven and two families of guards force every winning strategy to simulate common transitions. Existing transformations remove the restriction on the rabbit's initial positions and yield PSPACE-completeness for the directed game with one hunter, even on oriented graphs with positive minimum outdegree. Adjusting the guard sizes also proves PSPACE-completeness of the firefighting game, in which the rabbit may remain at its current vertex.

Video understanding benchmarks are the primary yardstick for measuring progress in video-language modeling, and a growing body of work relies on them to claim improvements. The reported benchmark scores in these works, however, are almost always produced by frameworks (e.g., lmms eval) used as black boxes: certain evaluation choices like frame sampling, preprocessing, and input formatting are inherited from preset defaults and seldom inspected carefully. We show these unexamined choices are a major source of variance. Even for the same model checkpoint, evaluation configuration alone can shift accuracy by up to 19.5pp. Hence, it is unclear whether reported gains reflect genuine improvements in capability or differences in evaluation setup. Moreover, many benchmark samples offer little evidence of visual capability, since they can be answered without the video, and the full suites are expensive to evaluate. We address these issues in two parts. First, we curate existing suites for discriminability and efficiency, producing Meta-Video, a balanced 1,000-question benchmark over 791 videos spanning diverse tasks and durations, which largely preserves model rankings, at 5 times inference speedup for a full-frame count sweep experiment (4-128 frames). Second, we establish a standardized evaluation protocol by sweeping the configuration space to isolate choices that measurably affect results, distill them into best practices and reporting standards, and release reproducible reference scores across five models and seven benchmarks. We hope that the proposed evaluation framework and the Meta-Video benchmark will support video understanding research by providing a reproducible and efficient foundation to measure progress in a reliable way.

Flow Matching has recently become popular as a technique for generative modeling with applications in many domains. In the standard Flow Matching formulation, we interpret the vector field as the solution of a least-squares regression problem, which is known to be sensitive to contamination. Therefore we leverage the viewpoint of robust statistics and replace the squared Euclidean distance with the Euclidean distance as a regression objective. This seemingly minor change strongly affects the properties of Flow Matching. We provide a probabilistic interpretation of the new regression target as a conditional geometric median and use this characterization to study distributional robustness of the Euclidean Distance-based Flow Matching. We analyze the mathematical properties of this changed objective and the effect of a contaminated target distribution on the involved conditional velocity distributions. We discuss the challenge of directly characterizing the distribution that our method induces. In specific contamination settings we show theoretically that our method yields an improvement for severe contamination. Finally, we present numerical experiments conducted on the CIFAR-10 image dataset where a proportion of the samples has been contaminated. Our Euclidean distance-based Flow Matching shows a clear advantage in scenarios of severe contamination of the target distribution.

Function-as-a-Service (FaaS) increases developer productivity by removing operational concerns such as managing hardware and software runtimes. Because function instances are started on demand, some requests must wait for a new instance, called a cold start. These are a common obstacle to more widespread adoption of FaaS, as they lead to unpredictable latency spikes that are not acceptable for some use cases. To solve this, many cold start optimizations have been implemented. However, comparatively little attention has been directed towards the slowest warm starts. With improvements in cold start times, slow warm starts emerge as a new problem for extreme tail events. By analyzing a dataset of several million requests across seven FaaS platforms, we show that the long tail of slow warm starts can be hundreds of times slower than their median performance, while a virtual machine we use as a baseline is substantially more consistent. We confirm this across four independent public datasets spanning 2020 to 2025, so the effect is neither new nor specific to one platform. Moreover, the warm share among the slowest calls starts to rise again at the highest percentiles: the most extreme latencies are not only cold starts, and cold-start mitigation alone cannot remove them.