Papers

In astrophysical complex TAIGA there are four Imaging Atmospheric Cherenkov telescopes (IACTs), located at a distance of 300-500 m from each other, due to which extensive air showers (EAS) generated by cosmic and gamma radiation are recorded by a single telescope or by multiple telescopes simultaneously (stereo-mode). The analysis of EAS images from two or more Cherenkov telescopes together is of particular interest, since for this approach the reconstruction of physical parameters of the primary gamma ray is much more accurate than in mono-observations. However, the gamma-ray event reconstruction strongly depends on the quality of the hadron background suppression of the classifier, therefore, the gamma event selection stage is one of the important ones in the further analysis of IACT data. We propose and study several neural network methods for gamma event selection in joint TAIGA-IACT image analysis. Neural network methods we suggested involves a single-stage analysis of IACT images from two telescopes, while the second method considers the formation of autoencoder-derived essential features that contain the EAS physical parameters, including the type of primary particle.

Test-time-training (TTT) layers update fast-weight memory while processing an input sequence. A gradient-based update, however, does not by itself establish how the memory revises an existing association. We study the structure and empirical role of prediction-derived updates in released TTT implementations. For dot-product objectives, we characterize the absence of a prediction-error residual and establish conditions under which inner-loop ascent and descent are equivalent after reparameterization. For the LayerNorm-based squared-error objective used by TTT-Linear and TTT-MLP, we derive the inner gradient and identify a prediction-derived component that cancels at the released initialization in the zero-epsilon limit. We then measure the full normalized prediction's contribution on three trained checkpoints. Removing this contribution while retaining the LayerNorm Jacobian gives a mean gradient norm ratio of 0.25 to 0.32 across the models. On 100,303 tokens from War and Peace, the same intervention with local per-token norm matching raises perplexity by 2.8-2.9% on TTT-Linear-760M and 1.3B, whereas freezing all fast-weight updates raises it by factors of 4.2-4.5. Separate synthetic tests illustrate the dependence of plain delta-rule overwriting on step size. The results distinguish a prediction-derived update from its behavioral value and show why implementation analysis and live interventions are both needed to assess test-time memory.

In this PhD thesis, we take an axiomatic approach to studying multivariate divergences, which are measures of the information content of a tuple of classical or quantum states. These divergences are studied in the context of large-sample and catalytic convertibility of tuples of states. This type of convertibility is characterized by the ordering of an axiomatically defined family of divergences. In this thesis, we identify this family of divergences explicitly for two scenarios, one classical and one quantum. In the classical setting, we show that the relevant divergences are certain generalizations of the bivariate Rényi divergences. Unlike previous studies, we do not require the distributions to have the same support, yielding interesting new divergences. In the quantum setting, we consider pairs of classical-quantum states with pure component states, and find that majorization is characterized by all $\alpha$-$z$ relative entropies for $\alpha<1$. This demonstrates the first operational interpretation of these relative entropies.

Conway's $99$-graph is a strongly regular graph with parameters $(v,k,\lambda,\mu) = (99,14,1,2)$ and whose existence has long been an open problem. We use both a sum-of-squares program and the fact that a strongly regular graph with $\lambda = 1, \mu = 2$ has no diamond or $K_{2,3}$ subgraph to derive a loose upper bound for the number of triangles that such a graph could contain. This bound is less than the number of triangles that the graph's spectrum asserts, implying that it does not exist.

Electroencephalography (EEG) is a non-invasive, scalable source of quantitative biomarkers for the early diagnosis of brain disorders, and deep representation learning has become a powerful paradigm for extracting information-rich latent features from raw signals. However, neural networks tend to under-represent high-frequency information, a concern that is especially acute for EEG given its oscillatory nature and characteristic $1/f^\chi$ power decay, which risks masking disease-relevant high-frequency signatures. In this study, we systematically investigate the spectral constraints of unsupervised time-domain feature extraction across three autoencoder architectures (LSTMVAE, VAEEG, Transformer) and a state-of-the-art EEG foundation model. Using synthetic experiments and empirical validation, we show that while autoencoders capture macro-scale shifts such as alpha slowing, they remain biased against higher frequencies due to a magnitude-driven masking effect induced by the Mean Squared Error loss under 1/f-decaying spectra. Increasing latent dimensionality mitigates this attenuation but sacrifices latent compactness and generative stability. We argue that, as the field advances toward large-scale foundation-level feature extractors, it is imperative to move beyond simple time-domain reconstruction and to integrate spectral-domain loss terms and gradient-balancing techniques, so that the subtle high-frequency transients critical for diagnosing complex brain disorders are preserved in the latent representation.

Mamba-based super-resolution models are efficient alternatives to Transformer-based restoration backbones, but their low-bit post-training quantization remains challenging because selective scan, Mamba's input-dependent state-space operator, induces different activation ranges across spatial tokens. Thus, a single layer-wise clipping range poorly matches all tokens. We propose Token-Group Quantization, a hardware-friendly activation PTQ framework that statically approximates a hardware-unfriendly post-scan grouping oracle, where tokens with similar preferred clipping ranges are assigned to the same group. To avoid runtime post-scan grouping in static low-bit inference, a lightweight predictor infers the groups from pre-scan features available before the quantized scan path. During calibration, we form pseudo-labels by clustering post-scan activation descriptors that capture similar rounding–clipping behavior. At inference, the frozen predictor maps pre-scan features to groups and selects entries from offline-calibrated group-wise clipping-bound tables, without runtime statistics collection, dynamic clipping updates, or post-scan group assignment. To address low-bit restoration degradation, we further introduce a Frequency-Preserving Refinement (FPR) objective for clipping-bound calibration. FPR aligns Fourier amplitude and phase between full-precision teacher and quantized student outputs to preserve high-frequency structures. Across PTQ protocols, bit-widths, backbones, and tasks, our method consistently outperforms prior Mamba and SR PTQ baselines, transfers to Transformer SR and Vision Mamba classification, composes with rotation-based quantization, and achieves 2.23x speedup on Jetson Orin Nano over FP16. Code is available here.

Current ASR models perform well on general benchmarks but still struggle with acoustically ambiguous words. Visual context can aid recognition in screen-based scenarios, yet existing benchmarks provide limited insight into the benefits of different types of visual evidence. To address this gap, we introduce ScreenASR-Bench , a tiered benchmark with three progressively challenging levels: Explicit Textual Evidence (ETE), based on direct on-screen text; Explicit Visual Evidence (EVE), using visible content rather than on-screen text; and Implicit Semantic Evidence (ISE), requiring scene-level inference. Extensive experiments on ASR models, Large Audio-Language Models (LALMs) and Omni models show that incorporating visual context substantially improves recall over the audio-only setting across all three model families. The gains are largest for ETE and smallest for ISE, indicating that models leverage direct visual evidence more effectively than indirect semantic cues. To facilitate future research on visually grounded speech recognition, we publicly release ScreenASR-Bench.

A graph is elementary if its edges can be coloured with two colours so that the two edges of every induced path on three vertices receive different colours. Elementary graphs are one of the two basic classes in the Chvátal–Sbihi decomposition of claw-free perfect graphs; they include all line graphs of bipartite multigraphs and all complements of bipartite graphs. We prove that every elementary graph $G$ satisfies $\ch(G)=\chi(G)=\omega(G)$, confirming for this class the conjecture of Gravier and Maffray that claw-free graphs are chromatic-choosable. The proof combines the Maffray–Reed description of elementary graphs as augmentations of line graphs of bipartite multigraphs, Galvin's theorem, an extremal (edge-maximal) choice of counterexample, and a Hall-type lemma describing how list colourings of a complete graph minus a matching can fail.

Head-mounted body and bimanual motion capture supports self-avatars, teleoperation, and embodied demonstrations. However, a single egocentric camera is vulnerable to a limited field of view, scale ambiguity, and hand-object occlusion. We present EgoWholeView, a wearable capture system of six synchronized wide-angle RGB cameras with a role-specialized pipeline. The outward views localize the rig in a world frame, while an overlapping downward pair independently predicts SMPL-X body and MANO hand parameters. Both predictors are adapted to egocentric fisheye data. To combine the two, reliability cues measured from the predictions themselves weight the fusion of body and hand joint rotations, while shape is aggregated across views and frames. The pair's baseline fixes metric scale, and triangulation recovers root translation. Relative to monocular prediction, fusion reduces PA-MPJPE by 3.7% for the body and 11.2% for the hands, and temporal refinement lowers angular-acceleration magnitude by 32.2% and 31.0%, respectively. On our six-camera recording, the pipeline produces body and hand states in a shared world frame using benchmark-fitted gates without retuning, and temporal refinement lowers hand acceleration magnitude by 41.08%. Using only RGB cameras and no third-person view, EgoWholeView offers a low-cost, portable route to egocentric interaction data.

Four-level indium-gallium-zinc oxide (IGZO) 2T0C memory needs reliable state formation, sensing references, and a path from cell errors to digital weights. We analyze 107 measured DC transfer curves from 37 isolated IGZO transistors (W/L = 3/0.5 micrometers). Four gate-bias operating points inferred from the first sweep are evaluated on later sweeps of 35 devices. Device-resolved operating points reduce second-sweep across-device current standard deviation by 49.4-52.4%, yet retaining initial thresholds increases misclassified points from 37/140 to 55/140. Leave-one-device-out reference estimation gives 0/140 on the second sweep; freezing second-sweep thresholds yields 6/140 on the third. Sampling eight reference transistors on the third sweep gives a mean held-out error fraction of 0.15% over 2,000 selections, versus 4.33% with frozen thresholds on matched held-out devices. We present conditional confusion matrices, a counterfactual common-gain sensitivity, and an INT4 error-propagation proxy. A simultaneous four-state bank with eight reference cells per state requires at least 32 physical reference cells, 12.5% of a proposed 256-cell data bank before peripheral overhead. A charge-conservation model and a partitioned hierarchical error model specify the pulse, retention, and array measurements needed to test transfer from DC device statistics to pulsed 2T0C operation.

Whole-body human-object interaction (HOI) generation requires coordinating body and finger motion, object trajectories, and time-varying contacts. Existing methods typically rely on iterative diffusion or test-time optimization, resulting in high inference costs. Direct one-step generation is highly desired in downstream applications due to efficient inference. However, it is challenging because these HOI components differ substantially in structure and scale. We present MFlowHOI, an efficient one-step framework for whole-body HOI generation. A Contact-Aware Multipart RVQVAE organizes body, hand, object, and contact signals into a structured latent space, with a shared decoder preserving their global coordination. An ST-DiT then models temporal and cross-part dependencies in this space and generates a complete interaction in a single MeanFlow evaluation. Predicted object-local contact correspondences further enable geometric refinement and interaction-preserving transfer to photorealistic 3D Gaussian avatars with different body proportions. Experiments demonstrate that MFlowHOI outperforms existing base generators in motion quality and text alignment while substantially reducing inference time.

Deep-search agents are commonly trained by distilling tool-use trajectories from stronger teacher models. However, structurally complex questions may remain relatively easy for teachers familiar with key entities in the underlying entity graph, resulting in lower search demand and weaker supervision. We refer to this teacher-specific variation in question difficulty as teacher-relative difficulty and investigate entity familiarity as an early predictor. We first introduce Evidence-Verified Entity Familiarity (EVF), which estimates familiarity by verifying a teacher-generated factual brief against online evidence. Across multiple teachers, EVF is significantly correlated with teacher-relative difficulty with AUCs of 0.60–0.70. However, its high cost makes it impractical for constructing entity graphs at scale. To address this limitation, we propose Self-Reported Entity Familiarity (SEF), a lightweight gating signal that directly asks a teacher whether it is familiar with an entity before admitting that entity into the graph. Across multiple teachers, trajectories from SEF-gated synthesis exhibit greater observed search demand and yield higher downstream student accuracy than those from ungated synthesis. Because familiarity is teacher specific, we further study cross-teacher reuse. We screen graphs collected for other teachers using Kimi-K3's SEF judgments and use the retained graphs to produce 5,411 additional trajectories. Combined with 2,200 trajectories from newly constructed, SEF-gated graphs, these data improve Qwen3.8-27B's accuracy from 44.10% to 66.41% on a BrowseComp subset, and achieving state-of-the-art performance among models of comparable size on BrowseComp, SEAL-0, HLE-text, and xbench. Overall, SEF provides a low-cost way to synthesize more informative supervision for deep-search distillation.

We revisit the problem of approximating a linear time-invariant LTI model by a reduced-order model (ROM) in an $H_2$-optimal sense. A widely used method for computing such ROMs is the iterative rational Krylov algorithm (IRKA). However, in practice, the stability of the resulting ROM is not generally guaranteed. In this work, we develop a novel framework that can be used to obtain $H_2$-optimal models of single-input single-output LTI systems while enforcing stability of the ROM. We use Malmquist-Takenaka orthogonal expansions in the corresponding $H_2$-spaces and their connection to hyperbolic geometry to develop an equilibrium function. We show that the poles of an $H_2$-optimal ROM transfer function are defined by the zeros of this equilibrium function. The resulting nonlinear root-finding problem is solved using a Newton-type method formulated on the products of Poincaré disk manifolds, where updates are performed along hyperbolic geodesics ensuring that the stability constraint is satisfied at every iteration. We present numerical experiments on benchmark systems for both discrete- and continuous-time LTI MOR problems, including examples where IRKA produces unstable reduced models while the proposed method preserves stability throughout the iteration.

We study sampling from $\pi(x)\propto\mathrm{e}^{-V(x)}$ on $\mathbb{R}^d$, where $V$ is $\alpha$-strongly convex and $\beta$-smooth with condition number $\kappa=\beta/\alpha$. For this problem, the proximal sampler preserves the target exactly and recovers overdamped Langevin dynamics in the small-step limit, but its Gaussian resampling step discards directional information. Underdamped Langevin dynamics retains such information through momentum to accelerate exploration, motivating a proximal construction that combines target preservation with directional persistence. We introduce reflected proximal dynamics, which combines proximal sampling with an exact reflection of the auxiliary variable, conditional kinetic motion, and controlled refreshment. We develop two realizations: a flux-resampled underdamped half-turn and a persistent-momentum quarter-turn. From a warm start with constant $\chi^2$ divergence, both methods achieve total variation error $\varepsilon$ using $\widetilde O(\sqrt{\kappa}d^{1/4})$ expected oracle queries. We further remove the supplied warm-start assumption by combining either sampler with existing posterior samplers in a proximal initialization stage with an optimized coarse step size. Accounting for both initialization and sampling yields $\widetilde O(\kappa^{6/7}d^{1/4})$ expected oracle queries. In both regimes, the dependence on the target accuracy $\varepsilon$ is $\mathrm{polylog}(1/\varepsilon)$, making reflected proximal dynamics a promising framework for accelerated high-accuracy log-concave sampling.

Video generation is moving from the cloud to phones for privacy and offline use, and applications often ask for several clips at a time, so on-device video generation is a throughput problem before it is a latency problem. Contemporary mobile SoCs carry both a GPU and a neural processing unit (NPU) on shared DRAM, yet existing on-device video diffusion systems run the denoiser on one engine while the other sits idle or only decodes the finished latent afterwards. Heterogeneous designs built for mobile LLMs do not carry over: they synchronise the two engines at every operator and partition weight tensors between them instead of sharing one stored copy. We first measure the platform limits that decide what can run on this class of SoC: how large one buffer may be, whether both runtimes can map the same buffer, and how much activation memory the NPU toolchain lets one graph hold. Guided by these measurements, we store the base weights once in INT8 in the layout the NPU reads natively; the GPU consumes the same bytes directly as integer dot-product operands, so neither runtime rearranges weights on load. We then flatten the denoising trajectory into its full sequence of block executions and cut it once at a block boundary, so the GPU runs the prefix and the NPU the suffix. The cut is placed from measured rather than modelled rates, including the decoder that only the NPU can run, and taken while both engines run together because each engine's speed changes when the other is active. We present HeteroVDiff, which serves a multi-expert video diffusion model across the NPU and GPU with 48% less weight memory than giving each engine its own copy. Generating videos back to back on a Galaxy S25 with decoding inside the pipeline, HeteroVDiff sustains 1.35x the stream throughput of NPU-only execution, and on VBench-2.0 it matches FP32 in the total score over all 18 dimensions.

World-action models should preserve the intended manipulation across equivalent descriptions of a spatial goal. We evaluate this consistency in 864 rollouts from three models across four RoboLab scenes, holding initial states and task goals fixed. Describing relations through the reference object rather than the object-to-move lowers the rate of reaching a stable placement by 12.2 percentage points on average. The effect of instruction rephrasing varies across spatial relations, so aggregate scores can hide sensitivity to wording. Standalone tests of the policies' language understanding identify relation reversals and show that adding images can disrupt otherwise correct text-only interpretations. Rollout analysis also distinguishes spatial placement from instruction following: moving the wrong object can satisfy the requested relation, and rates of stable placement fall from 43.2% at any time to 21.8% at episode end. These findings motivate a more thorough evaluation of world-action models' steerability, language understanding, and grounding. Evaluations should test consistency across equivalent instructions, verify that the intended object moved, and check whether the requested arrangement persists to the end of the rollout.

The Recursive Projection Aggregation (RPA) decoder for Reed-Muller codes uses, in its default implementation, all one-dimensional subspaces at each level of recursion. We restrict our attention to the case where the number of errors is $t<d_{\mathrm{min}}/2$. It is known that if the number of errors is $t<d_{\mathrm{min}}/2$, then $2t+1$ subspaces at each level of recursion are sufficient to correct these errors. We show that, for structured error patterns, the number of subspaces required for correction can be reduced further by appropriately selecting the subspaces used for projection. In particular, when errors occur as a single burst of length $t < d_{\mathrm{min}}/2$, three subspaces at each level of recursion suffice for correction. We obtain similar results for multiple bursts and other structured error patterns as well.

LLM agents increasingly control robots using VLA policies, perception models, and procedural knowledge. These modules are difficult to reuse because their interfaces, embodiment assumptions, and conditions of use differ, and their benefits on physical tasks remain unclear. We package them as embodied skills that extend agent skills with explicit embodiment assumptions and conditions of use, building 250 action, perception, and experience skills from 1,052 resources. On this basis we introduce EmbodiedSkillBench, which runs an agent in a closed observation-action loop and records success, actions, tokens, and wall-clock time. It supports three modes: native control, skill-only execution, and Agent-Directed Reuse (ADR), a mode we propose in which the agent keeps its native controls and decides for each skill whether to invoke, modify, or ignore it. Across four agents on 100 perturbed LIBERO-Pro tasks, we find that skills help an agent only as far as they exceed what it can already do. Skill-only execution lifts GPT-5.6-Sol from 11% to 63% and Kimi-K3 and Claude-Opus-5 by about 20 points, while cutting their token usage by 58% to 85%. Yet all four agents then end up between 61% and 64%, and GPT-6-Astra falls from 92% under native control to 61%. This suggests that as agents grow stronger, the role of embodied skills shifts from supplying missing capability to reducing cost: GPT-6-Astra already solves most tasks through native control, so restricting it to skills removes more than it adds. Under ADR, which lets it keep native control and call skills selectively, GPT-6-Astra recovers to 85%, within 7 points of native control, while using 14.1% fewer tokens and 11.9% less time. With Jev as a skill choosing motion directions and gripper actions, it reaches 81% with 50% fewer tokens and 57% less time. The code is available at https://github.com/HanZhi306/EmbodiedSkill-Bench.

Morphogenesis is driven by the interplay of genetic regulation and tissue mechanics. In vivo video data is limited, particularly under genetic mutations and environmental perturbations, making generalization beyond the most representative conditions in the dataset challenging. However, at tissue scales, the dynamics of Drosophila can be described as a continuum flow, motivating the usage of representations learned across other physical systems as priors for developmental dynamics. We investigate finetuning Walrus, a Continuum Dynamics Foundation Model pretrained across a broad set of simulated continuous systems, on tissue velocity fields from only 12 wild-type Drosophila blastoderm videos. We evaluate autoregressive prediction both in-distribution and zero-shot under unseen temperature and genetic perturbations. Zero-shot refers to evaluation on perturbation conditions never seen during wild-type finetuning, without any additional training. In-distribution, a Fourier Neural Operator model trained from scratch achieves best performance on VRMSE. However, Walrus (CRPS) is the strongest model under the most severe distribution shift of unseen samples of mutated embryos and temperature conditions. Interestingly, predictions capture the shape of the mean surface tissue velocity trajectory and its separation between 17{\deg}C and 27{\deg}C, including the timing of the peak at Germ Band Extension onset, a biologically meaningful milestone of embryo development. Our results provide preliminary support that representations learned from different dynamical systems can transfer to experimental morphogenesis in developmental regimes, provided they admit a continuum description. More broadly, our results motivate further study of physics foundation models as a new complementary tool for the simulation of morphogenesis in low-data regimes.

Visual grounding is a central challenge for GUI agents, requiring an instruction to be mapped to a target within a crowded screenshot. Coordinate generation with single-point supervision reduces a spatially extended target to one position, despite the presence of many valid click locations. Region-based patch scoring alleviates this ambiguity, but its spatial decisions remain tied to visual-token footprints that may span the target, neighboring controls, and background. To address these limitations, we propose Seg2Click, which reformulates GUI grounding as executable-region prediction: identifying where a click can validly execute the given instruction. Using bounding-box supervision, Seg2Click predicts an executable-region mask and computes the final click point through a deterministic rule. This formulation captures the spatial extent of valid click locations and enables finer localization beyond the visual-token grid, making it well suited to small, densely arranged controls without requiring a verifier or zoom-in pass. Experiments across five backbone configurations from two model families demonstrate the effectiveness of Seg2Click. Seg2Click-9B achieves 92.06%, 94.10%, and 55.22% on ScreenSpot, ScreenSpot-v2, and ScreenSpot-Pro, respectively, while matched-backbone comparisons show improvements over both coordinate generation and patch-based grounding. Code will be available at https://github.com/jcwang0602/Seg2Click.

Understanding the relationship between generalization and hallucination requires clarifying the role of memorization. We study this relationship from an information-theoretic perspective through a Selector-Density Conditional Mutual Information (CMI) framework. In the sparse random-fact limit, a uniformly bounded information budget per fact forces a nonvanishing false-positive rate at recall bounded away from zero, even when all facts are observed during training. In a rules-and-facts setting, we decompose model-data mutual information into rule learning, atomic-fact memory, and excess memorization, revealing competition between rule learning and factual memory even without excess memorization. A joint budget bound quantifies the resulting constraint on atomic hallucination under perfect atomic recall. Moreover, an additive Gaussian model shows that maintaining high recall for atomic and composite facts incurs false positives even under optimal data mixing. Finally, controlled relational experiments hint how training exposure, rule complexity, and model capacity shape these trade-offs.

In-silico optimization of nucleic acids relies on predictive oracles that imperfectly capture biological function, so sequences optimized against one model may exploit model-specific artifacts and fail to generalize to independent models. Recent studies have shown that not all designers generalize equally well to independent models. In this work, we use agentic program search to iteratively propose, execute, evaluate, and revise code for nucleic-acid designers. We describe the successful strategies used to prompt the search. AutoNovo designers achieved the best observed mean among the evaluated methods on 3/3 DNA-design tasks in terms of in-silico performance, and 6/6 transfer tasks in terms of independent models. The designers come from a family of discovered nucleic acid designers rather than a single universal best designer. We plan to release the resulting designers. These results are the first steps in creating nucleic acid designers that maximize cross-model generalization, and thus improving the odds that in-silico designed sequences reflect real biology.

As large language models are increasingly adopted as multi-party communication facilitators, there is a growing demand for agents capable of balancing diverse group dynamics across contexts. However, existing LLM facilitators are often fine-tuned for task-specific objectives; generalized LLM-based facilitation methods that can adapt to different types of tasks and diverse group dynamics are underdeveloped. To address this gap, we present GUIDE (Generalized Unified Intervention for Dynamic Exchanges, a framework for developing generalized LLM-based facilitators across different multi-party contexts. GUIDE evaluates facilitation across both objective task outcomes and subjective metrics, including psychological safety and perceived friction. Within this framework, we study whether LLMs can self-learn generalized facilitation strategy offline to real-time interventions. We first collect multi-party human interactions guided by task-specific expert LLM facilitators (N=432 conversations). We then employ a self-learning method in which an LLM induces an actionable strategic playbook from conversations and iteratively refines it using feedback from an LLM judge aligned with human annotators (N=20), developing a single generalized facilitator that outperforms the three task-specific expert facilitators on held-out conversations 61.4% win rate. Finally, we evaluate these improvements empirically; in a real-time online empirical study, the self-learned generalized facilitator preserves facilitator performance and group productivity, improves speaking comfort and reduces interpersonal friction, while using 24.7% fewer turns and 29.0% fewer words than the expert prompts. These findings provide empirical evidence that LLMs can learn generalizable, task-agnostic social facilitation from task-specific human interactions, on-par with hand-crafted task-specific prompts across diverse group settings.

Harnessing rich modality features has recently gained prominence in multi-modal recommendation systems (MRSs). To delineate user modality preferences, existing works focus on modality fusion, which can be achieved via attentive summation or concatenation. In this work, we introduce the modality inhibitance effect - a phenomenon which limits existing MRSs when these models transit from dual-modal to tri-modal configurations resulting in significant performance degradation. In particular, apart from the modal-specific contamination encapsulated in the newly introduced modality, the inclusion of new features results in excessive informational redundancy, thereby severely impeding effective integration within current approaches. To address the modality inhibitance effect, we propose a new multi-scale modality wavelet (MMWave) multi-modal recommender, which aims to fully capitalize on essential modality features and suppress modal-specific noise. This is acheived by employing wavelet decomposition from a multi-scale (time & frequency) perspective, ensuring that only essential (non-redundant) features of each modality are extracted. We then account for the latent associations between item modality features by designing a new structure-aware multi-modal graph learning module that captures high-order latent structural information pertaining to items. To modulate the relative contributions of modality features and achieve effective integration, we then propose a new adaptive modality regulator component that employs a new cross-weight integration strategy. Experiments conducted on four real-world datasets validate the efficacy of our proposed model. The source code for this work has been made publicly available at https://github.com/kennethorq/MMWave.

Network decision-making tasks, including adaptive bitrate streaming, load balancing, and congestion control, require policies to process heterogeneous states and actions while balancing near- and long-term consequences. Existing solutions generalize poorly across network tasks, require substantial task-specific redesign, and lack reliable evaluation of long- horizon action consequences. Methods that learn through online interaction additionally face unstable optimization. We present NetRL, a two-stage LLM-based reinforcement learning framework that combines supervised fine-tuning on demonstration trajectories with online policy optimization. A universal mixed-embedding interface combines textual task descriptions with precise numerical states, while a network action head produces valid decisions. At its core, NetGRPO evaluates candidate actions over multiple future trajectories, directs deeper search toward uncertain alternatives through confidence-aware budgeting, and performs group-relative, clipped, and KL-regularized updates. NetRL achieves 23.6% higher QoE than NetLLM, 12.2% lower job completion time than Genet, and 24.1% higher network utility than Orca, yielding state-of-the-art performance among the evaluated methods. Confidence-aware budgeting reduces end-to-end training time by 27.1% without reducing validation QoE. NetRL also converges stably and generalizes to unseen network traces without retraining. Code and model weights are available at the repository linked in this paper.

Text-to-image models have made remarkable progress in visual fidelity and instruction following, yet complex compositional generation remains challenging. Existing agentic approaches address this limitation through iterative generation and editing, adapting their actions based on interaction history. However, they typically fix the continuation state (i.e., the image from which refinement proceeds) to the latest image. Since iterative editing is not necessarily monotonic, the agent may continue refining from the latest image even when it has already degraded. In this work, we introduce GenRevisit, an agentic framework that extends action selection to joint continuation-state and action selection. The agent can not only use interaction history to adjust editing instructions but also select an image candidate to edit, jointly deciding where to continue from and how to refine. We train GenRevisit with SFT followed by agentic reinforcement learning. To address the coarse credit assignment of final outcome reward in GRPO, which assigns the same trajectory-level advantage to all actions, we further propose Normalized Action Credit Assignment (NACA) to redistribute positive advantage according to each action's contribution to reducing the remaining objective. On GenEval2, GenRevisit achieves GM scores of 84.3 and 74.4 with Qwen-Image-2512 and Z-Image-Turbo, surpassing the strongest baselines using the same generators by 12.7 and 10.9 points respectively. GenRevisit achieves state-of-the-art performance on GenEval, GenEval++, and GenEval2 among baselines.

Commercial balancing wheels are closed systems: an internal controller keeps the board upright, and a rider steers by shifting weight and twisting the hips. We augment a one-wheel with a lead-screw weight shift for longitudinal motion and a reaction wheel for yaw. The reaction wheel stores limited momentum, so a turn is a sequence of torque pulses, and whether a pulse rotates the body depends on static friction at the tire contact. From a 24-pulse torque sweep on one surface we fit a two-part model that gives the probability that a pulse rotates the body, treating the breakaway torque as random, and the rotation angle when it does; at maximum torque the posterior engagement probability is 0.54 (90% credible interval 0.24-0.84). The weight shift is modeled as a rate- limited input with a measured speed gain and an assumed first-order response, which predicts stopping distances four times those of a direct-force model. These models drive an event-triggered supervisor that schedules pulses by expected heading improvement, brakes when the predicted braking path would leave the safe set, tightens margins for model uncertainty, and updates the friction estimate online. In simulation on 500 plants drawn from the fitted models, the robust supervisor completed 92.6% of missions with 1.8% safety violations and 1.6% safe aborts, versus 84.0% and 7.2% for a supervisor using mean models and 1.8% and 20.4% for a baseline that ignores friction. A video of the hardware experiments is available online (https://github.com/ASAS-Labs/x-wheel).

Participant-disjoint cross-validation prevents direct participant overlap, but it may still produce optimistic estimates when acquisition site and diagnosis are structurally dependent. We examined Parkinson's disease classification using 159 pressure-walkway summary features from 133 WearGait-PD participants, including 65 participants with Parkinson's disease and 68 controls. The cohort was non-factorial: two sites contained only one diagnostic group, whereas VA-Seattle was the only site containing both groups. In repeated participant-level nested cross-validation, site alone achieved an area under the receiver operating characteristic curve (AUC) of 0.955 plus or minus 0.003, exceeding the gait-feature model (0.882 plus or minus 0.014). Under the evaluated logistic-regression pipelines, adding gait features to site, age, and sex produced no measurable improvement in pooled discrimination. More site-controlled evaluations produced lower and imprecise estimates: VA-Seattle same-site AUC was 0.668 (95% bootstrap confidence interval 0.477 to 0.836), and the JHOC + Bayview-to-VA-Seattle transport stress test achieved 0.576 (0.375 to 0.770). Supporting within-diagnosis site-prediction and site-conditioned permutation analyses showed that the gait representation retained substantial acquisition-site structure. These findings indicate that participant-disjoint pooled validation did not isolate disease-related discrimination from site-associated variation. They encourage careful examination of site-label dependence and, where possible, support for pooled results through within-site or independent validation before broader claims are made about generalization across institutions.

Time series (TS) anomaly detection (AD) lacks a universal detector, with the best choice determined by the series rather than by the algorithm. Existing methods therefore either select one detector for each series or average a pool of them. However, both assume that no label ever arrives, while a deployed monitor does receive them, since an operator confirms the alarms it raises. In this paper, we propose ONESHOT, which uses a single confirmed anomaly to weight the pool rather than to choose one detector from it. Specifically, we score each detector on the confirmed anomaly, and those scores become the weights of a soft mixture over the whole pool. We show that a single confirmed anomaly is enough to come close to the best detector for the series. Across 152 univariate and 73 multivariate series, ONESHOT outperforms the strongest single detector, averaging the whole pool, and even picking the best detector on the confirmed event. Code is available at: https://github.com/seunghan96/one_shot.

Embedded platforms increasingly host safety-related streaming functions next to less critical ones. Whether a safety architecture pattern, such as a doer/checker pair, protects a function depends on the cores that run its components. Their schedule must meet the throughput requirement, and mixing safety integrity levels (SILs) on one core raises development cost. This report presents SafeDSE, a design space exploration framework for synchronous dataflow applications. One declarative constraint model decides the pattern of each actor, the placement that the fault model requires, the static-order schedule, and the platform, and prices SIL promotion. Guards activate the chosen patterns in a fixed graph that holds all of them, and node potentials encode throughput. An independent verifier re-checks every solution, and a generator turns its checks into evidence of a safety argument in Goal Structuring Notation. On 9 benchmark instances under period bounds, no rule that fixes patterns before mapping was both always feasible and never more expensive under conservative pricing, while joint exploration was. In the generated arguments, 49 to 62% of the leaves cite verifier checks; the rest are claims a mapping cannot discharge. Optimality within the model was proven for all synthetic graphs with up to 16 actors. This extended version adds the full model, the verifier and generator in detail, the complete catalog and parameters, and a proof of the potential characterization. It also reports every experiment per instance, a mutation analysis of the verifier, and the negative results of the development.