Papers

Reconfigurable Intelligent Surfaces (RIS) are emerging as a key technology for programmable wireless environments in the beyond the fifth generation (B5G) networks. However, data-driven RIS research remains bottleneck by the lack of standardized, high-fidelity and open-source datasets. In this paper, we introduce a large-scale 3GPP TR 38.901-compliant dataset for RIS-aided millimeter wave (mmWave) networks, that considers severe path loss, blockage sensitivity, and spatial channel sparsity make the RIS assistance more impactful. The dataset spans various canonical 3GPP deployment scenarios across 20 controlled variants, capturing diverse user densities, fading conditions, and blockage regimes. Uniquely, every sample includes oracle RIS phase configurations obtained via a globally optimal brute-force codebook search, providing gold-standard supervision labels that are absent from any existing public dataset. Rich multi-task annotations comprising full channel state information (CSI), per-link channel decomposition, optimal phase matrices, and channel quality index (CQI) labels support a broad range of machine learning paradigms and downstream tasks, including phase optimization, channel estimation, and interference management. As the primary benchmark task, we introduce a novel CSI-to-CQI mapping that frames RIS-aided link-quality prediction as a scalable scalar classification problem, thereby avoiding the exponential output complexity of the direct phase vector prediction. We have evaluated this mapping against state-of-the-art architectures under in-distribution, out-of-distribution, and real-world hardware measurement conditions. Our dataset provides a reproducible, extensible, and community-ready foundation to accelerate data-driven research in RIS-aided B5G networks.

The extreme environment within neutron stars presents the opportunity to probe the nuclear equation of state at high densities while studying the properties of these stellar remnants. Historically, the equation of state and neutron star mass distribution have been inferred separately, with the latter often simply assumed to be fixed in studies of the former, but the dependence of the equation of state on neutron star mass indicates they should be inferred together. In this work, we extend the generalized version of the popular RIFT algorithm known as Hyperpipe to interface with flexible, user-provided priors that will facilitate joint inference of equation of state and binary neutron star population hyperparameters. We demonstrate this framework's utility via application to a combination of several existing observations, particularly including massive galactic pulsars, double neutron star systems, millisecond X-ray pulsars, the gravitational wave event GW170817 and the nuclear symmetry energy, within the context of a widely-adopted parametric EOS family and a simple Gaussian population model. We recover parameters consistent with previous analyses of the EOS and binary neutron star population, finding the latter to fit a bivariate normal distribution with mean $(\mu_1,\mu_2) = (1.39,1.27) M_{\odot}$ and width $\sigma = 0.08 M_{\odot}$. Furthermore, we implement novel coordinate transformations in our pipeline, with which we have discovered that the ad hoc prior boundaries used for this EOS family may be too restrictive. We present results with relaxed yet still physical prior boundaries, noting vastly improved hyperparameter posteriors and modestly different equation of state inferences.

World-action models (WAMs) jointly generate future world states and actions through iterative denoising, using shared weights to process heterogeneous semantic streams of video, proprioceptive, and action tokens. Quantization reduces inference cost, but comparable numerical errors in different streams can have markedly different effects on final actions, making numerical accuracy alone insufficient for reliable control. We introduce SteerQuant, a 4-bit quantization framework for WAMs that steers errors toward computations with less influence on final actions. It maps how each stream's quantization errors affect final actions and uses this map to guide shared channel scaling. Activation scaling is further calibrated for each stream and denoising step to accommodate changes in activation ranges and action impact. This adapts quantization to different stream requirements without duplicating weights or increasing bit-widths for selected streams. To reduce the extra kernel launches and memory traffic introduced by scaling, we develop Rudder, a 4-bit inference engine for WAMs that fuses scaling and output compensation into low-bit kernels. Under W4A8 and W4A4, SteerQuant maintains mean LIBERO success within 0.8 percentage points of full precision, while delivering up to $2.23\times$ denoising speedup over BF16 across three WAMs with reduced peak GPU memory usage. On a real dual-arm robot, W4A8 deployment achieves a $1.35\times$ end-to-end inference speedup while maintaining average task success relative to BF16.

How can we personalize a shared expert library from a user's pairwise choices? Prior work can realize different reward trade-offs by merging reward-specialized experts, given a vector of trade-off weights. In practice, users can more naturally choose between outputs than specify numerical weights. The challenge is therefore to turn these choices into the coefficients required for merging, while accounting for ambiguity when feedback is limited. Our key idea is to treat the unknown reward weights as latent variables: infer a posterior over them from pairwise choices and reward-score differences, and use its mean directly as the merge coefficients. We instantiate this idea as Bayesian Pairwise Merging (BPM), whose posterior also characterizes which reward trade-offs remain plausible given the feedback. We evaluate BPM on radiology summarization, image captioning, and story generation, spanning text-to-text and image-to-text generation. With 100 feedback per simulated persona, BPM achieves macro decided win rates of 91.7%, 77.1%, and 64.3% against uniform merge. For six pairs of simulated personas, each prefers the model fitted to its own feedback, a pattern also observed in a human proof-of-concept. In simulations under BPM's model and prior, its nominal 90% intervals for temperature-scaled reward weights achieve task-averaged marginal coverage of 88.9% and 89.2% with only 10 and 25 comparisons, respectively. BPM thus enables personalization from pairwise feedback without per-user policy training, while characterizing the coefficient ambiguity left by limited feedback.

In municipal planning workshops, planners and other stakeholders compare shared-micromobility parking policies by varying no-parking zones, retained sites, spacing, or area allocations. Each edit changes feasible sites and how much demand they cover, so the alternative must be reoptimized on the same spatial data. Full-set greedy, the transparent reference for this task, takes tens of seconds per alternative at city scale. We present Constraint-exact Low-latency Iterative Planning with Pooled Evaluation and Replay (CLIPPER), an optimizer with audit functions developed for requirements elicited with the City of Braunschweig. In each greedy round, it forms a deterministic candidate pool of bounded size, computes how much still-uncovered demand each candidate would add, and rejects candidates that violate an active constraint. An optional offline audit scans every remaining feasible candidate and records what the restricted pool omitted. We evaluate these functions on complete eleven-state edit chains ($E_0,\ldots,E_{10}$) in Braunschweig, Munich, and Berlin. With $K=1024$ candidates per group, the fixed-width mode CLIPPER-F has mean coverage gaps to full-set greedy under the same policy of 0.245, 0.003, and 0.001 percentage points in Braunschweig, Munich, and Berlin, respectively, while mean rollout time falls by factors of 13.6–28.9; no audited run terminates while a candidate outside the pool could still increase coverage. Plans computed from two checksummed versions of Braunschweig's official no-parking-zone data differ in 30 of about 540 selected sites although coverage moves by only about 0.1 percentage points. These changes still require municipal assessment and implementation. The findings inform a proposed municipal process that versions policy inputs, reports site changes beside coverage, and scans the full candidate set before a final decision.

Accurate long-time prediction of Cahn-Hilliard dynamics, particularly in the late-stage coarsening regime, is computationally demanding because it requires simulations over extended physical time. Various machine learning approaches have been developed to speed up the simulations through efficient surrogate evaluations. While the learned map can be accurate over a single learned interval, small prediction error will accumulate when the map is repeatedly applied in a long-time prediction, leading to large numerical errors. In this work, we identify the structure function for the coarsening dynamics of the Cahn-Hilliard equation and propose a two-stage structure-preserving Fourier Neural Operator (FNO) learning framework. Stage~I introduces a bound-conforming FNO (bcFNO) with a soft, magnitude-dependent amplitude penalty. Stage~II retains the bound-conforming objective and supplements it with structure-function regularization, yielding a structure-preserving FNO (spFNO) that corrects late-stage coarsening statistics. Numerical experiments on the various spatial resolutions and reference-solver time steps suggest that the bcFNO stabilizes the long-time prediction by suppressing large amplitude excursions, without changing the low online cost of the FNO. Furthermore, the structure-function term consistently reduces discrepancies in the normalized structure function and provides further correction in associated late-stage coarsening quantities, including the $L^3(t)$ growth trend and the scaling collapse. These results indicate that our multi-stage implementation of physical principles in operator learning could be applied to other multiscale systems with statistical scaling, such as the functionalized Cahn–Hilliard equations and turbulence, provided that the state constraints and statistical observables are adapted to the governing dynamics.

We study the online transportation problem, in which $n$ requests arriving sequentially in a metric space must be irrevocably assigned to $k$ capacitated facilities. Beyond classical logistics applications, this problem models resource-allocation tasks arising in machine learning, including online facility assignments, recommender systems, and mixture-of-experts routing. We introduce Robustified Greedy (RG), a deterministic generalization of the Robust Matching algorithm that achieves a competitive ratio of $6.6604k-2.89$, improving upon the state-of-the-art bounds of $8k-7$ (Arndt et al., SOSA 2026) and $8k-5$ (Harada and Itoh, ICALP 2025). RG also retains the metric-sensitive guarantee established for Robust Matching (RM) (Nayyar and Raghvendra, FOCS 2017), achieving a competitive ratio of $O(k^{1-1/d}\log^2 n)$ in $d$-dimensional Euclidean spaces for fixed $d>1$. No comparable metric-sensitive guarantee is known for the transportation algorithms of Arndt et al.\ or Harada and Itoh. Beyond these competitive guarantees, RG provides a simple explanation for its decisions. It favors the natural nearest-neighbor assignment and, for suitable parameters, departs from this choice only when it identifies a reassignment that reduces the cost of its maintained auxiliary matching, thereby correcting accumulated assignment costs. We also prove that nearest-neighbor assignments account for a guaranteed fraction of RG's total cost, approaching one-half for appropriate parameters, even under adversarial arrivals. Experiments on real-world datasets corroborate the theory: RG achieves lower cost-to-Opt ratios than the competing algorithms while retaining a substantial nearest-neighbor component in its cost.

Existing multimodal video highlight detectors typically assume that visual, audio, and textual streams are continuously available. In practice, however, inputs may suffer from localized frame missingness or complete-stream outage. We formulate this robustness challenge along two dimensions: temporal missingness, where frames are missing independently in each modality, and stream-level missingness, where one modality is unavailable throughout a video. Moreover, we find that the mean squared error (MSE) loss is misaligned with both the evaluation metrics and the peak-driven nature of highlights. Therefore, we propose Temporal-Stream Modality Dropout (TSMD), which combines structured missingness simulation with a joint objective comprising pointwise MSE, per-video Pearson correlation, and peak-oriented RankNet loss terms. TSMD has three variants: temporal, stream-level, and mixed dropout. On the MoSu and Mr. HiSum datasets, TSMD-Temporal improves mAP@15 by 7.06 and 3.41 points over TripleSumm under 50% independent temporal removal, whereas TSMD-Stream performs the best under complete-stream removal. TSMD-Mix retains most of these complementary benefits and ranks the best or the second-best across the evaluated temporal and stream-level conditions.

As multi-agent systems enter high-stakes domains, the possibility that agents may circumvent safety boundaries is a growing concern. Prior work has examined this risk primarily in adversarial settings, where agents are instructed or rewarded to communicate covertly and evade oversight. We show that benign agents can cross the same boundaries without adversarial incentives. We emulate a software-engineering workflow in which a planner represents a company hiring an external developer. The planner writes requirements and holds a company credential it is instructed not to disclose to the developer; a monitor screens their exchanges. Seven of nine tested frontier models disguise the credential in their requirements to help the developer recover it while evading the monitor, even after completing their assigned objective. For example, across 6,000 episodes with DeepSeek-V4-Pro, the planner attempts concealment in 16.9%; in 0.9%, the credential evades the monitor and is recovered and used by the developer. Despite the low per-episode rate, the risk compounds across repeated exchanges: at 0.9%, 105 independent episodes carry a 61.3% chance of at least one breach. Our analysis suggests that the planner misinterprets disguised disclosure as helping the developer and reads the nondisclosure rule as banning plaintext, not character codes or riddles. We call this behavior covert assistance. Monitoring favors concealment, while concealment gives way to direct disclosure when the recipient is presented as a human rather than an AI agent, so leakage persists. Giving the monitor the credential improves detection but requires trusting it with the secret. These risks, in models already used for software engineering, challenge oversight to distinguish authorized cooperation from task-advancing assistance that crosses safety boundaries.

A large language model (LLM) can rate depression severity directly from a social media post, or mark which clinical criteria the post shows and let a simple rule turn the count into a label. The latter is easier to audit because a clinician can check each marked criterion. We compare these approaches on two Reddit corpora using three LLMs (from 9B to frontier scale) and two questionnaires (PHQ-9, BDI-II), and measure agreement with quadratic weighted kappa. For the two frontier models, criteria extraction scores above chain-of-thought on one corpus only when its decision thresholds are fitted on labeled data. Neither model's gain is significant, with or without recalibrating chain-of-thought on the same labels. With thresholds fixed a priori from PHQ-9's criteria, extraction shows no gain on either corpus, even where models mark over two criteria per post. The 9B model behaves differently on a corpus from depression communities. It labels most posts as severe, whether prompted directly or with chain-of-thought, while the a priori rule, which uses no labels, beats both. After chain-of-thought is recalibrated on the same labels, no significant gap remains, consistent with a calibration effect. Yet higher ordinal agreement does not ensure better detection of severe cases. PHQ-9 criteria extraction misses most severe posts, and moving from direct prompting to chain-of-thought and then to extraction increases misses in nearly all comparisons. On one corpus, a relabeled stress dataset, a model using that dataset's own features, some computed from the post text, is not significantly different from frontier criteria extraction under the a priori rule.

Protein optimization remains a longstanding goal in life sciences. Existing machine learning-assisted directed evolution (MLDE) methods primarily rely on sequence-only features, overlooking the critical spatial constraints and co-evolutionary interactions encoded in protein structures. However, directly integrating structural information remains challenging due to the scarcity of reliable mutant structures. To address these issues, we propose StructEvo, a novel structure-aware reinforcement learning framework for protein directed evolution. StructEvo employs a delta-structure fusion encoder to approximate mutant structure features via feature differences, enabling dynamic incorporation of spatial knowledge. The vast mutation space is then decomposed into manageable subspaces through a structure-aligned hierarchical action network, while a geometric constraint further stabilizes delta feature learning. Our approach outperforms prior state-of-the-art methods by 9.2% and 16.3% on two challenging optimization benchmarks, and further identifies an experimentally validated epistasis pattern in GFP, highlighting the importance of structural guidance for effective protein directed evolution.

Interactive video generation (IVG) models have achieved remarkable progress in producing controllable visual content guided by user-defined actions, yet their security vulnerabilities remain largely unexplored. In this paper, we present the first systematic study of backdoor attacks against the interactivity of IVG models. Based on this attack surface, we propose BadAction, which leverages action-guided triggers to achieve the attack. Specifically, BadAction implants predefined motion patterns into the action sequences of backdoor samples and associates them with a static target video. Once triggered, the backdoored model generates frozen future frames that no longer respond to subsequent user actions, while preserving normal behavior on benign action sequences. In addition, we explore a stealthier attack in which multimodal triggers jointly poison action, text, and image inputs. Experiments show that BadAction achieves average attack success rates of 91.0% with action-only triggers and 80.4% with multimodal triggers. Moreover, extensive defense evaluations show that BadAction successfully bypasses existing backdoor detection methods, revealing a critical security gap in the interactive video generation pipeline. Project page: https://wsad55.github.io/badaction01/.

Nonlinear exceptional points (NEPs) were proposed to circumvent the noise–responsivity cancellation that limits linear exceptional-point sensors. At an NEP, the stationary frequency response can be singular while the Hamiltonian Petermann factor remains finite, suggesting a genuine signal-to-noise ratio (SNR) enhancement. Yet recent model-specific analyses of nonlinear fluctuations have reached conflicting conclusions. We prove that any singular stationary frequency response requires a fixed-point bifurcation of the nonlinear dynamics. In local fluctuation regimes, the singularities in responsivity and long-time frequency uncertainty are both set by the critical relaxation rate, leading to exact cancellation in the SNR and precluding any divergent enhancement. Furthermore, we derive a path-information bound that applies across both local and nonlinear fluctuation regimes and places a finite ceiling on the long-time SNR enhancement. Notably, this upper bound is attainable in the local fluctuation regime, implying that frequency readout under such parameter settings is information-optimal.

Recent advances in large language models have made automatic game generation increasingly feasible, yet reliably improving generated games beyond a playable version remains challenging. Naive iterative refinement can easily overfit a small set of test cases, producing fragile games with unresolved bugs, missing behaviors, and poor generalization to broader player interactions. We introduce RSIGame, an autonomous agentic game development framework with recursive self-improvement. RSIGame organizes development into complementary local and global loops. Concretely, a local explore-diagnose-improve loop broadly explores the executable game, diagnoses and prioritizes discovered issues, and performs evidence-grounded revision, where an evolving checklist continually accumulates new testing and improvement guidance. A global loop tracks overall quality, preserves the best checkpoint, and detects saturation or regression over long-horizon development. Beyond test-time improvement, RSIGame further internalizes successful development experience into the generator through training. Across 140 GameCraft-Bench tasks, two game engines, and five generators, RSIGame consistently improves game quality under matched development budgets. Notably, experience internalization enables Qwen3.8-27B to reach 61.38 on Godot and 58.53 on Phaser, exceeding GPT-5.5 one-shot scores while reducing Qwen's generation tokens by 11 times.

Creating sound effects for a new game-character skin requires a distinct acoustic identity while preserving gameplay-event roles. The challenge is to complete a coherent set of related sounds whose required degrees of redesign differ. We formulate this task as completion conditioned on base-skin audio, completed target assets, and a textual design description. We develop a pipeline to collect, process, and align corresponding events across League of Legends skins. Building on Stable Audio 3's pretrained audio prior, we fine-tune a latent inpainting model to jointly complete missing events. A signed soft retention mask encodes available audio and an adjustable transformation hint for each missing event, specifying the requested balance between retention and redesign. Experiments on held-out skins show improved reconstruction over the evaluated general-purpose audio editors. Target-derived hints further improve paired similarity, with three-level hints retaining most of the benefit of continuous guidance.

Large language models (LLMs) enable rich semantic user profiles for recommendation, but such profiles are more expensive to generate and are not necessarily desirable to deploy uniformly. We study whether LLM-generated profiles can instead be invoked selectively within a production recommendation pipeline. Using a real-world streaming dataset covering movies, TV shows, and sports content, we train a serving-time routing gate that assigns each user to either a collaborative sequential recommendation model or a recommendation model driven by an LLM-generated profile. The gate uses only serving-time features and learns to identify users for whom profile-based routing can increase Novelty@10 while preserving ranking relevance. A routing threshold controls how aggressively users are sent to the generative model, exposing a tunable novelty–relevance trade-off. At an overall NDCG-loss budget of 5%, the learned gate increases Novelty@10 by 6.5% while routing 12.5% of users, outperforming simple heuristic and random routing policies at comparable relevance cost. These results show that LLM-generated user profiles can serve as a controllable complement to collaborative recommendation, while results with non-generative semantic profiles indicate that the benefit stems from selective routing rather than LLM generation alone.

Broadening the study of quantum optimization algorithms from binary to $k$-element alphabets has been shown to open new avenues for potential quantum advantage. A quantum advantage claim for approximate optimization requires showing that, under the same assumptions, an efficient quantum algorithm provably achieves a better performance than can be proven for the best known efficient classical algorithms. We study local classical and quantum algorithms for Max-$k$-Cut on $d$-regular graphs of girth $g$. We advance classical algorithms for this problem by developing a local vector algorithm based on the explicit vector construction of Thompson, Parekh, and Marwaha (TPM). Our algorithm gives the best provable cut fraction guarantee among known efficient classical algorithms on regular graphs for $k\geq3$. To evaluate the performance of QAOA under identical assumptions of girth and regularity, we develop tensor network techniques for general $d$ and $k \geq 2$. In addition, using an equivalence to a coupled qudit–boson system, we compute the QAOA performance in the infinite-degree limit. Together, these techniques give provable guarantees on QAOA performance on large graphs. Despite the improvements we introduce to the classical algorithm, QAOA achieves a better cut fraction guarantee for depths $p\geq 9$, corresponding to girth $g\geq 20$, for both finite- and infinite-degree regimes. Thus, we obtain an apparent quantum advantage from applying QAOA to the Max-$k$-Cut problem.

Partially miscible ganglia trapped within porous media, spanning one or multiple pores and evolving through diffusive mass transfer, are common in subsurface (e.g., CO$_2$ and H$_2$ storage) and manufacturing (e.g., fuel cells) applications. We present the ganglion network model (GNM), a reduced-order method for simulating how a population of such ganglia evolves inside an arbitrary porous microstructure. GNM operates on a tree graph, the ganglion network, extracted from the pore-scale image of a given sample. Each point on the graph encodes a possible ganglion configuration in the void space, without any loss of geometric or topological complexity. The evolution of a population is modeled by representing each ganglion as a particle on the graph and tracking it according to a set of rules formulated herein. The rules capture capillary events such as pore invasion, retraction, snap-off, fragmentation, and merger. Unlike pore-network models, another graph-based modeling tool at the pore scale, GNM solves no system of equations and its cost scales with ganglion count, not domain size. We validate GNM against an image-based pore-network model in 2.5D and 3D domains with populations undergoing ripening, dissolution, and growth. We find good agreement in ganglion statistics, aggregate properties, and spatial configuration. We further argue that the ganglion network is the statistical space needed for extending kinetic theories of Ostwald ripening from single- to multi-pore ganglia, and provide an outline for how to do this. GNM opens the door to modeling other dynamics of trapped phases in porous media.

We present a new measurement of the stellar and dark matter distributions in massive early-type galaxies at redshift $0.4<z<0.6$ by combining stacked strong and weak lensing measurements. The stacked weak lensing measurements down to the small radius of $0.013\,\mathrm{Mpc}/h$ enabled by the Subaru Telescope Hyper Suprime-Cam data are combined with stacked enclosed projected mass measurements from 13 strong lens systems. We find that adding strong lensing constraints improves constraints on the stellar and dark matter distributions. The central dark matter profile is found to be consistent with the standard Navarro-Frenk-White density profile. The stellar initial mass function is found to be bottom-heavy. The total density profile in the inner region is described well by the power-law density profile with the slope of $-1.96\pm 0.07$, which can also explain weak lensing signals out to $1\,\mathrm{Mpc}/h$ reasonably well. The stacked weak lensing profile constrains an internal mass-sheet transformation parameter to $\lambda_{\mathrm{c}}=0.978\pm0.010$ assuming a single cored component. Our analysis demonstrates that the combination of the stacked strong and weak lensing serves as a powerful tool for studying central density profiles of galaxies.

Recent DESI observations hint at a preference for dynamical dark energy with an equation of state (EoS) that crosses the phantom divide at $w=-1$. In this work, we investigate metastable dark energy (DE) on the minimal phantom brane. Braneworld effects lead to $w < -1$ at early times, whereas metastable decay of DE results in $w > -1$ at late times. We consider three metastable decay prescriptions on the phantom brane: an exponentially decaying DE component (M1), decay of DE into non-baryonic dark matter (M2), and decay into dark radiation (M3). Using CMB observations, DESI DR2 BAO measurements, and the DES-Dovekie Type Ia supernova sample, we compare these models with $\Lambda$CDM, the Chevallier–Polarski–Linder (CPL) parametrization, and the minimal phantom-brane model with a non-decaying $\Lambda$. We find that metastable models on the brane are preferred over $\Lambda$CDM, their decay rates being $\Gamma/H_0=0.362\pm0.086$, $0.539\pm0.125$, and $0.483\pm0.112$ for M1, M2, and M3, respectively. Best fits of metastable models improve over $\Lambda$CDM by $\Delta\chi^2_{\rm MAP}=-17.14$, $-18.76$, and $-18.22$, while the corresponding $\Delta{\rm DIC}$ values are $-8.66$, $-9.63$, and $-6.56$. Braneworld screening drives effective phantom behaviour at $z\gtrsim0.4$, while DE decay drives the EoS toward non-phantom values closer to the present epoch, with the phantom divide $w=-1$ being crossed at $z \simeq 0.3$–$0.4$. These results show that the interplay between metastable decay and phantom-brane screening provides a simple physical mechanism for an effective crossing of the phantom divide consistent with DESI observations.

Generative policies have emerged as a promising paradigm for robot learning, combining expressive generative action modeling with scalable imitation learning from large demonstration corpora. However, heterogeneous demonstrations can induce suboptimal action chunks whose errors compound over time, eventually driving the robot into out-of-distribution states from which recovery is difficult. Action verification offers a test-time scaling strategy for mitigating this failure mode by sampling multiple candidate actions and using a verifier to select one for execution. Existing approaches, however, remain temporally myopic and costly to train, evaluating candidates from the current observation alone without accounting for trajectory continuity and often relying on large verifiers and additional expert demonstrations. In this paper, we introduce Temporal Verification (TeV), an efficient temporally aware action verification framework for flow-matching VLAs. TeV first learns a temporal token that summarizes recent observation–action history, enabling candidate chunks to be evaluated as continuations of the execution trajectory rather than as isolated predictions. Conditioned on this token, TeV constructs positive–negative pairs without additional expert demonstrations or preference annotations and trains an energy-based verifier contrastively to assign lower energy to higher-quality, trajectory-consistent action chunks. Beyond post-hoc ranking, TeV further uses the learned energy landscape to guide intermediate flow samples toward lower-energy regions, improving candidates before final selection. Extensive experiments in simulation and real-world settings demonstrate that TeV provides reliably ranks action candidates, improves task success rates, and produces smoother execution trajectories.

Before a validator can be promoted to a hard gate on a deployment pipeline, it has to be shown that its firing separates outputs that reach users in working order from those that do not. We run that screen on 13 validators in a deployed generative agent, against 550 runtime and 350 static builds labelled by downstream outcome, and report each check's marginal separation $J=\mathrm{TPR}-\mathrm{FPR}$ with Newcombe intervals and Fisher exact tests. Two checks survive correction for multiple comparisons, two more are nominal only, and the remaining nine are not distinguishable from zero, three of them because they never fired on any sampled build. Execution itself is not random with respect to the property being gated, and this replicates: across four runs covering 1,867 builds and ten distinct runtime checks, probes were skipped on 144 of 895 broken builds and 1 of 972 acceptable builds (per-run rates 15.6% to 16.6% against at most 0.3%), every skip carrying the same unsafe-to-probe reason. Because a skipped check is recorded as a pass, this imposes a ceiling that no check quality can lift: a check that needs a live artifact cannot operationally detect more than about 84% of broken builds in this harness. For the one check with construct-specific labels, a detector built for blank output fires on 0 of 90 human-labelled blank builds (95% upper bound on sensitivity 3.3%), and the global frame statistic it approximates separates the classes only weakly (AUC 0.59), so the gap is not a threshold that needs tuning. The same gap appears one layer up: on a census of tens of thousands of judge-scored builds, 32.5% of rejections carry no recorded issue at all. We argue that evaluation records must distinguish a check that ran and passed from one that did not run, must carry the evidence for a rejection, and that an inventory of checks is not evidence about a gate.

The Dark Energy Spectroscopic Instrument (DESI) Main Survey Data Release 1 (DR1) provides an unprecedented spectroscopic view of galaxies and quasars across a large fraction of the observable Universe. However, the survey's tracer-dependent targeting strategy introduces complex observational selection effects that strongly influence the apparent distribution of galaxies in the redshift-luminosity plane. In this work, we investigate the origin of the characteristic "fork" geometry observed in DESI DR1 and show that the observed distribution is consistent with an interpretation in which it arises from the combined action of intrinsic galaxy bimodality and the survey's redshift-dependent selection function. We show that the observed underdensity separating the dominant galaxy populations is influenced not only by the physical Green Valley but also by survey selection effects, since it is further shaped by magnitude limits, surface-brightness selection, and structural incompleteness. Using DESI DR1 spectroscopy together with Legacy Survey photometry, we analyze the roles of the BGS, LRG, ELG, and QSO target classes in producing the observed distribution and assess their impact on luminosity-function measurements. Our results indicate that both target selection and observational biases significantly affect the observed galaxy distribution and the evolutionary trends inferred from it. We summarize practical observational considerations aimed at minimizing these effects and enabling more robust studies of galaxy evolution with DESI.

Passenger access doors are safety-critical subsystems in railway vehicles, yet detecting abnormal door behavior in real operation is challenging because faults are rare, diverse, and often unlabeled. This paper addresses railway door condition monitoring as a cycle-level unsupervised anomaly detection problem, where each complete opening-dwell-closing cycle is treated as a single monitoring unit. We propose the Temporal Cycle-Aware Attention Autoencoder with Cross-Signal Consistency (TCAA-CS), trained exclusively on nominal cycles. It combines a dual-stream encoder that processes continuous physical measurements (position, current, voltage) and binary logical states (door-closed, door-locked) through separate 1D-CNN branches, an LSTM encoder with temporal attention pooling, and a triple hybrid anomaly score fusing reconstruction error, latent-space deviation, and phase-aware cross-signal consistency. The consistency term helps identify cases where individual signals appear plausible but their inter-signal relationships become physically or logically inconsistent. On real industrial data from a passenger train in commercial service, TCAA-CS achieves 93.8% recall, 97.3% precision, and a 0.5% false-alarm rate, outperforming representative unsupervised baselines. System-level evaluation on an NVIDIA Jetson AGX Xavier supports the feasibility of real-time onboard deployment.

Designing expressive sequence layers with efficient inference remains a central challenge in modern machine learning. Standard softmax attention achieves excellent sequence modeling performance through rich nonlinear token interactions, but it requires a key-value cache that grows linearly with sequence length, limiting its scalability. Linear attention enables efficient recurrent computation with a constant memory footprint, yet its reduced expressivity often yields inferior modeling performance. We introduce Switching Linear Attention (SwiLA), a novel sequence layer that bridges this gap by enhancing representational capacity while retaining the fixed-size recurrent state of linear attention. We derive the SwiLA recurrence from the test-time regression framework, casting the state update rule as online expectation-maximization in a mixture of linear regressions model. At test time, each output dimension dynamically selects among multiple linear attention components based on the input. Across associative recall, in-context language learning, and language modeling benchmarks, SwiLA shows strong performance and narrows the gap to softmax attention, even surpassing it in several settings.

CLIP is a powerful vision-language model, but it was not designed for fine-grained defect localization; CLIP-based anomaly detectors therefore adapt it with prompts or lightweight modules to increase defect sensitivity. We show that stronger sensitivity does not necessarily make local evidence reliable: under domain shift, adapted CLIP-AD models often assign high anomaly scores to both true defects and visually complex normal regions. The issue is not simply missing defect information, but a local scoring rule that decodes defect and hard-normal evidence, having the same anomaly evidence. We propose TED (Text-Axis Evidence Decomposition), a post-hoc scoring method that asks whether each ambiguous response is better supported by source defect patches or by source normal patches mistaken as anomalous. TED compares these supports under the host's normal-versus-anomaly text response, leaves the backbone and prompts unchanged, and requires no target-domain training. It works as a train-free score for raw VLM backbones or as a source-calibrated residual correction for adapted CLIP-AD hosts. Across frozen VLM backbones, TED substantially improves pixel-level localization over raw prompt similarity; across adapted hosts, it improves most pixel-level settings over P-AUROC, P-PRO, and P-AP. Gains are largest under stronger hard-FP competition, with mean localization gain increasing from +5.0 in low-competition regimes to about +10.9 in mid/high-competition regimes. These results suggest that recoverable defect evidence can already exist in pretrained multimodal representations, but reliable localization requires decoding it against hard-normal competitors. Code will be released at TED GitHub repository.

We investigate whether predicted Mean Opinion Scores (MOS) can reliably support system-level comparisons of speech enhancement (SE) methods by introducing system-level preference accuracy (SPA). Although MOS prediction models are widely used to evaluate SE systems, their performance is typically assessed by correlation with human-rated MOS, which does not guarantee agreement on which system is better. SPA addresses this gap by directly evaluating whether predicted and human-rated MOS yield the same system preferences. Using SPA, we systematically evaluate three settings: single prediction models, ensembling, and domain adaptation. Through experiments, SPA varies substantially across single prediction models, from 9.4% to 76.8%. Even the best model disagrees with human judgments in approximately 23% of system comparisons. Ensembling yields only limited improvement, while domain adaptation tends to substantially improve SPA in the closed condition but brings only modest gains in the more practical open condition, where neither the target systems nor the speakers are known. These results suggest that SPA can reveal errors correlation-based evaluation alone does not expose, and that predicted MOS alone can lead to unreliable conclusions in practical SE system comparison.

We prove that equivalent locally compact, locally Hausdorff, étale groupoids with paracompact unit spaces have Morita equivalent reduced and full $L^p$-operator algebras for every $p\in[1,\infty]$. The reduced theorem requires pairing-valued approximate identities, reflecting the failure of unconditionality of the reduced norm for $1<p<\infty$, while the full theorem rests on a full-clopen reduction theorem proved by dilating spatial representations. For $1\leq p<\infty$, we also compare the concrete $p$-linking algebra of the reduced Morita equivalence with the reduced $L^p$-operator algebra of the linking groupoid: the canonical comparison is contractive, injective, and has dense range in general, and is an isometric isomorphism for $p=2$. We further study Morita cycles arising from proper étale groupoid correspondences, including the reduced case under suitable extension hypotheses. Applications include an $L^p$ version of Green's symmetric imprimitivity theorem, results for coarse groupoids and inverse semigroups, and invariance of the corresponding $K$-theory.

Audio and speech models are released rapidly, but reported scores often conflate model capability with deployment and evaluation choices. The same checkpoint can produce different predictions under different runtimes, hardware, decoding settings, or fallback policies. Even fixed predictions can receive different scores under different normalization and metric implementations. Existing speech benchmarks standardize selected datasets or scoring procedures, but rarely connect heterogeneous model onboarding, controlled inference, and versioned scoring in one executable workflow. We introduce SURE-EVAL, a Systematic and Unified Agentic framework for Reproducible Evaluation of audio and speech systems. A Tool Agent Workflow converts model releases into isolated, verified callable tools. A Main Agent Workflow commits task-specific inference and scoring protocols, then delegates all score-bearing operations to versioned deterministic programs. Each result retains its runtime, protocol, pipeline nodes, predictions, and audit artifacts. Across 18 public releases covering automatic speech recognition, text-to-speech, voice conversion, speaker diarization, speaker-attributed recognition, and multi-task audio understanding, a Codex-only baseline completes 12 models in one shot, while the same agent with the SURE-EVAL Tool Agent Workflow completes all 18. We also conduct unified evaluations over seven ASR test conditions and two TTS subsets. A protocol analysis of three TTS systems finds absolute differences of 0.02-0.52 points between paper-reported and unified results, with the direction varying by model and language. These results show that reproducible evaluation requires controlling both model execution and output scoring.

I investigate the threshold double logarithmic behavior of the fragmentation function of a gluon into $^1S_0^{[1,8]}$ quarkonia, which is suppressed by $1/N$ in Mellin space. While threshold double logarithms have been resummed to all orders in the strong coupling for gluon fragmentation functions for quarkonia at leading power in $1/N$, it has not been known how to resum them for channels suppressed by powers of $1/N$. Threshold resummation can be important for the $^1S_0^{[1,8]}$ channels, as the tree-level gluon fragmentation functions have finite yet nonzero values at the kinematical threshold. I derive a formalism for resumming the threshold double logarithms for these channels at the leading double logarithmic level and discuss the phenomenological impact.