Papers

Echocardiography is the most widely used cardiac imaging modality, yet interpretation demands integrating visual evidence across global anatomy, localized structures and dynamic cardiac motion. Machine-learning models have automated individual tasks, but they are typically built for a single purpose and depend on expensively labeled datasets - a barrier particularly acute in pediatric care, where data are scarce and anatomy changes with age. Here we present EchoDino, a self-supervised foundation model for echocardiography, created by adapting the DINOv3 framework to 3.7 million frames from 1.7 million unlabeled pediatric echocardiography videos. With its encoder frozen, EchoDino produces representations that capture global context, local anatomy, and dense spatial detail. We introduce Motion-biased Entropy Maximization Sampling (MEMS) to select the most informative frames for video-level analysis. Across nine pediatric and adult datasets, EchoDino outperformed strong baseline models, raising view-classification accuracy from 0.609 to 0.889 and the area under the receiver operating characteristic curve for structural-heart-disease detection from 0.811 to 0.872, while also cutting age-estimation error from 3.857 to 1.389 years, achieving the best segmentation accuracy and lowering ejection-fraction errors. By generalizing from label-free pediatric data to adult echocardiography, EchoDino offers a versatile foundation for cardiac image analysis across the lifespan.

We establish dimension bounds and an exact dimension formula for equilibrium measures of degree $d$ holomorphic endomorphisms of $\mathbb P^2=\mathbb P^2(\mathbb C)$ satisfying suitable expansion and domination conditions. Write $\lambda_1>\lambda_2>0$ for their Lyapunov exponents. If two inverse histories ending at the same point determine different fast directions, we prove that ${\mathrm dim}_H \mu_F>\log d/\lambda_1+\log d/\lambda_2$ when $\lambda_2>\log d$, and that ${\mathrm dim}_H \mu_F=2\log d/\lambda_2$ when $\lambda_2\ge2\log d$. Applying these results to an explicit quadratic family, we disprove the Binder–DeMarco conjecture on a non-empty open set of holomorphic endomorphisms. The proof adapts the projection growth strategy of Li–Pan–Tong–Xu to the holomorphic setting, combining strong leaf geometry and the Ledrappier–Young theory for endomorphisms with a multidimensional extension of Wu's restricted sum estimate.

Existing attacks on unlearned diffusion models assume that the erased concepts are known in advance and focus on recovering them. In practice, however, model providers may not disclose which concepts have been removed, and even with access to the original base model, an adversary may still lack a clear target to attack. In this paper, we aim to answer the following critical but overlooked questions: which concepts have been erased from the model, and how many have been erased in total? To this end, we present Tracer, a framework that rapidly and accurately identifies erased concepts and estimates their number. Tracer efficiently identifies erased concepts without generating and classifying images. By combining lightweight spectral analysis of weight footprints, it enables efficient search over large candidate vocabularies. To distinguish multiple erased concepts, we introduce a footprint coverage objective that guides sequential discovery. Tracer estimates the number of erased concepts by detecting a sharp decline in candidate confidence as the selected concepts account for the erasure footprint, without requiring labeled examples for calibration. The framework requires only lightweight linear algebra and limited forward probes, with no prior knowledge of the unlearning algorithm. Experiments across text-to-image and text-to-video backbones and diverse unlearning methods demonstrate that Tracer identifies erased concepts and estimates their number in seconds, achieving 150 to 137,000 times and 133 to 20,000 times speedups over MIA and brute-force search on image and video models, respectively, with substantially higher identification accuracy.

High-dimensional bioinformatics data, characterized by a large number of features relative to the number of samples, pose major challenges such as the “curse of dimensionality,” leading to overfitting, high computational cost, and poor generalization. Traditional feature selection methods often suffer from limited scalability and adaptability in such domains. We propose an LLM-in-the-loop reinforcement learning (RL) framework for bioinformatics feature selection, where the RL agent formulates feature selection as a sequential decision-making task, while the large language model (LLM) enhances the process in two ways: (1) guiding exploration through domain-informed advice, and (2) providing hybrid rewards that integrate data-driven performance with knowledge-driven evaluation. The LLM also produces explanations to improve interpretability for human experts without altering the RL policy update. Experiments on diverse bioinformatics datasets show that the LLM-in-the-loop framework outperforms baselines, achieves stable performance across downstream models, and converges faster than pure RL.

Temporal dependence can separate the Gaussian approximation of stochastic gradient descent from its stationary moments. For unmodified least-squares SGD, we construct a design with standard Gaussian marginals whose stationary error has every positive moment infinite. Independent observations with the same marginals instead give finite stationary variance. Both regimes retain a Gaussian small-step limit. Our general theory establishes pathwise contraction from a finite second design moment, then uses score cancellation and localization to obtain stationary Gaussian and Ornstein–Uhlenbeck limits. Independent Gaussian regression errors yield an exact conditional Gaussian law and total-variation convergence under the same design integrability. Stronger design conditions identify a positive first-order total-variation constant and a deterministic covariance correction with $o(a)$ error. A scalar coverage expansion translates this correction into its inference consequence. Experiments examine distributional error, coverage, and calibration with dependent scores. Together, these results establish precise probability-law approximation beyond moment-based stationary analysis.

An organization often holds too little labeled data to train a model that generalizes, and the records that would supply the rest sit with organizations that cannot release them. Cross-silo federated learning offers a way through, since participants exchange model parameters rather than records, but it ordinarily settles two aspects of the arrangement in advance, the participating sites and the classes the model can predict, and deployment can breach both. A new site joins after training, once the established sites have finished their engagement and gone offline, and its records arrive unlabeled, mixing conditions the model already recognizes with conditions no participant has observed. We present an autonomous three-stage procedure that expands the model entirely at the joining site: reconstruction experts screen for novelty, clustering separates the flagged records into candidate conditions, and class means describe the old classes, all inside one shared representation. Those classes were learned from records that never leave their owners, so the usual defenses against forgetting are unavailable, and the procedure supplies the evidence they would have carried from either of two dissimilar sources, prototypes held by the federation or records held by the joining site. On a real industrial condition-monitoring dataset, run end to end with no label consulted, either source holds old-class accuracy at 0.868 or above with forgetting at most 0.063, and the two differ by 0.021, so a configuration can be chosen by the disclosure it permits rather than the accuracy it delivers. Both keep old- and new-class accuracy in balance where every alternative we measure gives up one for the other, and both retain more of the old classes than distillation- and regularization-based baselines. The balance still holds with only 6 labeled records per arriving condition and 3 retained per old class.

Language models process and generate text sequentially in token units, and the tokenizer determines how much text each inference step covers. Under standard tokenization, a short English phrase such as "On the table." is usually produced as four separate predictions for the preposition (On), article (the), noun (table), and punctuation (.), where each consumes a sequence position and adds inference cost. We introduce CoBPE, a compositional tokenization approach that represents such phrases as a lexical base token (table) attached with a small set of reusable surface modifiers, composed in embedding space at input and predicted jointly at output. In controlled pretraining from scratch at 780M and 1.3B scales, CoBPE shortens sequences by 30% and improves average downstream performance by 1.2 points relative to standard BPE under matched training compute. Our results suggest that part of what is now expressed through token sequences can instead be modeled through structured representations, opening a broad design space for more token-efficient and capable language models.

Repeated impacts of a droplet on a vertically forced bath can produce a growing Faraday-wave response under forcing conditions for which the corresponding initially undisturbed bath remains linearly stable. We investigate this mechanism using axisymmetric two-phase direct numerical simulations over a wide range of Reynolds and Weber numbers. The resulting phenomenological regime map identifies quasi-periodic, decaying, faint, and unstable-bath responses. Along the parameter path considered here, the transition to the unstable-bath regime occurs over a narrow range of Weber numbers and varies only weakly with Reynolds number. In the unstable regime, the bath develops a coherent subharmonic response at half the forcing frequency, while disturbances in the decaying regime remain broadband and weaken between successive rebounds. Bath-only simulations show that a finite-amplitude perturbation representative of droplet impact can seed the confined radial mode susceptible to parametric forcing, initiating a growing subharmonic response below the linear instability threshold of the initially flat bath. A reduced capillary-dominated model provides a scaling estimate for the impact-equivalent perturbation amplitude, and bath-only simulations initialized with these perturbations reproduce the bounded-to-growing transition observed in the coupled simulations. Floquet analysis of the finite cylindrical bath identifies the first axisymmetric Neumann mode as having the lowest subharmonic instability threshold. A time-resolved Bessel decomposition independently shows that the growing bath response is dominated by the same radial mode, while the droplet-induced subharmonic response projects predominantly onto this structure. These results identify a finite-amplitude route by which localized droplet impacts can access Faraday-wave dynamics below the linear onset threshold of an initially undisturbed bath.

Local Gaussian models of constant-step learning predict output variability and expected losses, but weak convergence alone does not justify these moment predictions. We establish moment-accurate Gaussian mixtures by matching stationary energy with local Ornstein–Uhlenbeck limits, ruling out quadratic tail mass invisible to weak convergence. For step size $a$, the second-order Wasserstein error is $o(\sqrt a)$, uniformly over invariant laws, using each law's actual root weights. The assumptions combine confinement, descent, finitely many hyperbolic equilibria and root continuity with finite-variance innovations. The result yields observable covariances, expected objective gaps and first-order mean shifts, while allowing singular covariances, compatible saddles and weights without a limit. For additive noise given by a fixed invertible transform of independent standardized Student $t_3$ coordinates, symmetry gives an order-sharp $\sqrt a$ smooth-test bound. Numerical transport calculations demonstrate the value of root-specific covariances; controlled SGD studies assess observable predictions across step sizes, batch sizes and model geometries.

We present a likelihood analysis of observables in deeply virtual exclusive $\pi^o$ electro-production off the proton. We use experimental data for the cross section as well as the beam and target spin asymmetries measured in various longitudinal polarization configurations at Jefferson Lab Hall A and Hall B. Performing a local extraction, for each kinematic bin, we derive a joint likelihood for the various helicity amplitudes defining the different polarization configurations. We subsequently extract the Compton form factors that parameterize the underlying QCD scattering amplitudes in the chiral-odd sector. Markov chain Monte Carlo (MCMC) methods are employed to explore the full parameter space, extracting the parameter uncertainties, correlations and credible intervals. The joint likelihood analysis of the cross-section and asymmetry data constrains the parameter space indicating that a larger kinematic coverage and additional polarization configurations, e.g. in the transverse sector, will improve the precision of the parameters. We also report a determination of the proton tensor charge and tensor anomalous magnetic moment extracted in a Bayesian statistical framework using deeply virtual exclusive scattering data.

A many-body wavefunction exponential in R, where the square of R is the sum of the squares of the interparticle distances, is shown to be the solution of Schrodingers equation for a potential proportional to the reciprocal of R, for N particles. A set of translationally invariant rotational eigenstates of angular momentum is found.

Charging a quantum battery through a pumped charger can lock a substantial part of its stored energy inside the bipartite correlations that grow between the constituent cells, and this correlation ergotropy, being the energy that no local single-cell operation can reach and that only collective operations can recover, is the object of the present study. Following the charging trajectory of two coupled three-level cells, we find that the correlation ergotropy closely tracks the logarithmic negativity, which shows that the locked energy is carried by genuine bipartite entanglement rather than by coherent collective dynamics alone, and that an equal-strength coherent non-entangling drive produces no such locking at all. For a pure two-cell battery we establish an exact, basis-independent closed form in which the locked energy depends only on the sorted eigenvalues of the reduced cell and the cell energy levels, revealing that the locked fraction rises from zero at the separable corners of the Schmidt simplex to unity at the symmetric maximally entangled state and remains invariant under local unitaries. Decoherence and dephasing, which destroy the correlations, do not release this energy but erase it, so extraction must be collective and must take place while the correlations persist, while cold damping leaves the effect intact and thermal occupation suppresses it. These results give a resource-theoretic reading of correlations in quantum batteries, sharpen the line between genuine entanglement and coherent collective dynamics, and point to experimentally accessible conditions for correlated-charge and collective-extraction protocols.

As pretraining increasingly repeats data, every run faces three questions: how many epochs to take, how that number should change with model size, and whether anything besides the epoch count matters. We answer them by pricing a repeated token against two references: one epoch on the same data, which gives its value, and fresh data at equal compute, which gives its cost. Against fresh data, the cost of repetition follows a single variable, the number of extra epochs divided by the unique tokens per parameter. Against the same data, a second epoch is worth nearly as much as a fresh one, and repeated tokens fall to half the value of fresh ones after a critical epoch count that grows with the training budget per parameter but hardly with model size. With unique data fixed, the predicted compute-optimal run grows model size and epochs together until loss stops improving, near the critical epoch count. The same variable accounts for the direction of size trends that appear to conflict: larger models tolerate fewer epochs when the corpus is fixed, from about 15 at 127M to 4 at 2B parameters, but not when unique data grow with the model. Counts alone do not determine loss: at identical counts, replaying shards consecutively raises loss by up to 0.46~bits per byte, concentrating repeats on fewer samples also raises it, lower-entropy sources degrade faster with repetition, and re-tokenizing repeats helps only under heavy repetition. These results offer an empirical guide to pretraining when unique data, rather than compute, are the binding constraint.

Cold-start drug-drug interaction (DDI) prediction tests whether models can identify clinically significant interactions for drugs without training-time interaction history. Existing benchmarks mostly report aggregate edge-prediction scores, leaving a key evaluation question unanswered: when models receive molecular, textual, or knowledge-graph (KG) evidence, do they actually use the evidence that pharmacologically supports the interaction? We introduce ColdDDI, a reconstructible diagnostic benchmark built from DrugBank 5.1.13, with 1,900 approved small-molecule drugs and 565,731 positive DDI pairs. ColdDDI evaluates pairs with zero, one, or two unseen drugs. It also annotates each interaction by whether it changes drug exposure or drug effect, and by whether the biomedical knowledge graph contains shared enzymes, transporters, or targets that can plausibly mediate the interaction. These annotations separate evidence availability from predictive dependence. We evaluate eight conventional DDI methods and 13 LLMs; for open-weight LLMs, we test five prompt patterns and use masking, drug replacement, and channel-sensitivity metrics to probe knowledge utilization. ColdDDI exposes that, in the hardest split where both drugs are unseen, the main performance divide is mediator availability. A fine-tuned 1B LLM recovers 89-93% of interactions with a shared enzyme, transporter, or target, but only 40-62% without such a mediator. More importantly, KG-provided evidence is not always used; several KG-augmented baselines change little when the shared mediator is masked or disrupted, whereas fine-tuned LLMs respond strongly to this intervention. Thus, ColdDDI evaluates knowledge utilization rather than knowledge access alone, showing where cold-start DDI models rely on mechanistic evidence and where they fail despite receiving it. Code is available at https://github.com/0217ljh/ColdDDI-NeurIPS2026.

The study of quantum memories out of thermal equilibrium is motivated both by the need for memories that withstand realistic noise in fault-tolerant quantum computers and by the fundamental question of when quantum memory can emerge in autonomous many-body systems. We establish a general equivalence between CSS-based quantum memories subject to Pauli noise and classical memories via two associated probabilistic processes. In the asynchronous setting, where the quantum dynamics is generated by a Lindbladian, we identify a broad class of syndrome-based Lindbladians for which the quantum memory problem is equivalent to memory in continuous-time classical Markov dynamics. In the synchronous setting, we introduce syndrome-based probabilistic cellular automata (SPCAs) and establish a one-to-one correspondence between their classical memories and CSS-based quantum memories under local quantum channels. This framework provides a systematic route for deriving and constructing quantum memories from classical probabilistic processes, and recovers known results, including the fact that the 4D toric code is a quantum memory in the presence of thermal noise. As an application, we give the first rigorous proof that the 4D toric code with Toom's rule is a quantum memory in the presence of local Pauli noise, provided the error is below a constant threshold. We prove this both for the corresponding synchronous, PCA-type quantum dynamics and for a Floquet Lindbladian. More generally, we show that classical memory in an SPCA can be lifted to a quantum memory under Floquet Lindbladian dynamics; whether this can be done with a time-independent Lindbladian remains open. Our results provide a general framework for relating classical and quantum memory, shed light on the conditions for fully autonomous quantum memories, and may guide the construction of provable autonomous quantum memories in low dimensions.

We study sender-receiver games in which players commit, before the state is drawn, to payments contingent on the subsequently verified state and the receiver's action. Payments may be transferred or burned, serving as contractual guarantees or promised tips. We analyze equilibrium guarantees, welfare, and computational complexity. For arbitrary finite action sets, we characterize pessimistic utility guarantees under unilateral commitments. Sender-only and sender-first commitments have worst-equilibrium welfare equal to the best constant-action welfare. Simultaneous commitments can instead reduce welfare below every equilibrium of the original game. Neither simultaneous nor sender-first commitments always admit an efficient equilibrium. Receiver-first commitments always admit an efficiently computable welfare-maximizing equilibrium, but can also reduce worst-equilibrium welfare. For binary actions, we characterize their entire equilibrium payoff region. With binary actions, optimal receiver commitments are computable in linear time, whereas optimizing deterministic sender commitments is NP-hard. Privately randomized sender commitments admit an exact $O(n\log n)$ algorithm, with an optimal lottery over at most two schemes differing at only one state. This yields an additive approximation by deterministic commitments. For arbitrary finite action sets, optimal privately randomized sender payoffs equal Bayesian persuasion values: ordinary persuasion with burning alone, and persuasion with nonnegative compliance payments when transfers are allowed. These equivalences yield polynomial-time algorithms with polynomial-size commitment lotteries.

Individual-level wildlife identification often suffers from data scarcity, as varying observations of the same animal under diverse poses, viewpoints, and motions are rarely available. Image-to-video (I2V) generation offers a promising way to mitigate this limitation by synthesizing additional observations from a single reference image. However, existing I2V models mainly emphasize global layout, semantics, and motion, and therefore often fail to preserve fine-grained local appearance cues that distinguish one wildlife individual from another, such as fur texture, stripe boundaries, spot configurations, and contour transitions. We observe that these identity-critical cues are closely related to high-frequency information. To address this challenge, we propose WildIcon, a high-frequency-guided I2V framework for wildlife individual consistency. Specifically, WildIcon introduces a frequency-aware identity encoding branch that extracts individual-specific high-frequency cues from the reference image. Combined with isolated foreground information, the resulting identity tokens are then injected into cross-attention blocks as identity conditioning. Building on a frozen backbone with lightweight identity adaptation, WildIcon preserves fine-grained identity cues visible in the reference image while retaining the motion controllability and semantic fidelity of the base I2V model. In addition, to support the training and evaluation of wildlife individual-consistent I2V, we construct WildlifeVid, a wildlife-centric video dataset with high-quality, temporally coherent clips and individual-level identity labels. Experiments on I2V generation and downstream animal re-identification (ReID) show that WildIcon achieves stronger individual consistency than existing baselines, and that its filtered outputs can serve as useful candidate training augmentations for downstream ReID.

As Large Language Models (LLMs) gain tool use capabilities such as web search, they can retrieve and cross-reference public information, creating privacy risks beyond memorization. One manifestation is re-identification: linking an anonymized interview transcript to a named individual. Yet without ground-truth identities, the coverage of such attacks and the protection offered by a defense cannot be reliably measured. We introduce AgentDOXX, an evaluation suite of 822 synthetic interview transcripts grounded in public information about real individuals with known identities. We evaluate fifteen configurations of open-weight and proprietary models, isolating the effect of web search, and analyze their search trajectories to distinguish retrieval-driven from parametric identifications. Ground-truth identities reveal that re-identification risk is distributed across an agent's execution: retrieval and parametric recall both contribute, with open-weight models identifying 15-28% of transcripts without search; identification succeeds in over 88% of cases once the target appears in a retrieved result; entity masking leaves at least one attacker successful on 85.3% of a stratified sample; and privacy instructions suppress naming but not retrieval, with configurations scoring 0% accuracy yet retrieving the subject in up to 62% of transcripts. We further show that observed attack trajectories can provide supervision for localizing identifying spans, offering a path toward attack-informed anonymization.

We establish optimal quantum-classical separations in communication complexity for search problems. We introduce a total search problem called Pelagic Fourier Fishing and show that it admits an $n$-qubit quantum one-way protocol, whereas every randomized two-way protocol requires $\Omega(2^n)$ bits of communication. We then introduce a variant of this problem whose solutions can be verified in polynomial time. This variant also admits an $n$-qubit quantum one-way protocol, while every randomized one-way protocol requires $\Omega(2^n)$ bits of communication. We also construct a family of efficiently verifiable total search problems achieving an $n$ versus $\Omega_d(n^d)$ separation between quantum one-way and randomized one-way communication for every fixed $d \ge 2$. In the quantum protocol, Alice prepares her message using a single unitary from the $d$th level of the Clifford hierarchy, and Bob performs a Clifford measurement. This separation is asymptotically optimal under this restriction on Alice's message. For $d = 2$, Alice's message is a stabilizer state and Bob's measurement is Clifford, so the protocol uses no magic, yet achieves an optimal quadratic quantum advantage. Finally, we discuss how these separations can be adapted to near-term quantum advantage experiments in which the demonstrated advantage is both unconditional and efficiently verifiable.

Extra inference compute is usually spent on sampling more reasoning chains. We study where inside an existing chain an additional continuation should begin. We define expansion utility, the change in correctness from restarting a chain at a stored step, and measure it at every eligible step for nine models on six benchmarks (41 model and benchmark cells). Restart position matters: steps selected on one set of continuations beat uniform placement when scored on disjoint ones, in held-out audits on 5, 16, and 38 cells (+4.25 points [+2.51, +6.63] in a fresh five-cell audit). A fixed rule that restarts from the last eligible steps, always-last, is a strong baseline: our learned router beats uniform placement but shows no detected gain over it, and on DeepSeek-R1-Distill-Qwen-14B/MATH-500 always-last exceeds the exact self-consistency frontier at matched aggregate generated output by +0.052 [+0.008, +0.098], using 0.774x the aggregate generated output of four-sample self-consistency. Cross-fitted oracle selection still finds held-out headroom beyond declared positional classes, a target for future selectors. Finally, breaking step-label ties by earliest index flips the sign of a pointwise selector's gain over uniform placement in every seed of a five-seed diagnostic with four rollouts per step; randomized ties remove the bias.

Passive leader arms transmit an operator's motion without returning remote contact or load cues. TacGELLO uses the existing servos of a 3D-printed leader for tactile contact feedback at the gripper trigger and current-based load cues at the arm joints. AnySkin contact detection activates a trigger position hold, while changes in follower joint current adjust leader-servo output limits to command resistance during lifting. Across six UR3 sessions with one operator, sent-command tracking error was 0.15–0.36$^\circ$ RMSE, with estimated trajectory lag of 82–98\,ms. Tactile onset was logged in 60 of 62 segmented TacGELLO grasp intervals. Median commanded closure after tactile or width-based contact references was 0.7–2.1\,mm with TacGELLO and 13.0–13.9\,mm with the original interface.

We show that unlike $\textsf{PFA}$, $\textsf{MA}_{\omega_{1}}$ is consistent with the existence of $2^{\omega_{1}}$-many Countryman lines which pairwise contain no uncountable isomorphic or reverse-isomorphic suborders. This clarifies an issue regarding the claim that $\textsf{MA}_{\omega_{1}}$ implies that there is a finite basis for the Countryman lines. The family of Countryman lines comes from a family of $2^{\omega_{1}}$-many coherent trees which pairwise contain no club-isomorphic subtrees. To prove this, we first construct such a family of trees using $\diamondsuit$, and then we iterate ccc forcings while preserving this property.

Subset currents on free groups, on surface groups and, more generally, on word-hyperbolic groups, were introduced by Kapovich and Nagnibeda, and further studied by Sasaki. Subset currents extend ordinary geodesic currents and provide measure-theoretic generalizations of conjugacy classes of infinite quasiconvex subgroups. Their supports can be viewed as “branching laminations". For free groups and surface groups, we construct non-rational subset currents with genuinely branching supports that carry no other subset currents except scalar multiples. For every nonabelian finite rank free group and every closed hyperbolic surface group, we show that such a current can be chosen so that its support is uncountable, is disjoint from the locus of two-point boundary subsets, and consists entirely of Cantor subsets of the boundary. More generally, we prove that every nonempty metrizable Choquet simplex occurs as a compact convex base for the cone of currents carried by a minimal genuinely branching subset lamination. The same universality phenomenon extends, apart from the usual rational-current terminology, to every non-elementary torsion-free hyperbolic group. We also study finite-pattern complexity. We prove that if $R_L(n)$ counts the allowed round patterns of radius $n$ in a minimal subset lamination $L$, then $R_L(n)=O(n)$ forces the carried-current cone to be finite-dimensional, while infinite-dimensionality forces $R_L(n)/n\to\infty$; this universal threshold is sharp. We show that the coarse growth type of $R_L$ is independent of the chosen free basis and, more generally, of the chosen finite marked graph. We also show that a refined switch system recovers the exact dimension of the carried-current cone.

Controlling tail risk in randomized algorithms has received increasing attention. Randomized online algorithms can achieve a lower expected competitive ratio than the competitive ratio of the best deterministic online algorithm. However, individual executions may still incur poor outcomes. Tail constraints address this risk by limiting the probability that the incurred ratio exceeds a prescribed threshold. We investigate how such constraints affect the structure and performance of randomized online bidding algorithms. By revisiting the classical primal-dual lower-bound framework for online bidding, we find that restricting the growth between consecutive bids can yield a stronger lower bound on the expected competitive ratio. We then show that tail constraints with zero exceedance probability impose an asymptotic upper bound on the ratios between consecutive bids. Incorporating this structural property into the primal-dual framework allows us to derive a stronger lower bound. Moreover, the dual analysis suggests a randomized geometric bidding algorithm whose performance matches this bound. The insight from our primal-dual analysis also guides the design of algorithms for general tail constraints allowing positive exceedance probabilities. Motivated by the geometric growth suggested by the dual solution, we study randomized periodic bidding algorithms. We show that, for any fixed geometric base, the uniform offset distribution is optimal in terms of both the expected competitive ratio and tail-constraint feasibility. We then characterize the exceedance probability of these algorithms and use this characterization to choose a geometric growth factor that satisfies the prescribed tail constraints, establishing parameter-dependent upper bounds on the optimal expected competitive ratio.

Academic dishonesty is an urgent problem in computer science (CS) education, compounded by the rise of artificial intelligence (AI) and the prevalence of online learning in a post-COVID world. Self-efficacy and the impostor phenomenon are psychological constructs linked to academic dishonesty, yet little research has examined how they relate to how students perceive it. This study investigates the interplay between the impostor phenomenon, self-efficacy, and CS students' perceptions of academic integrity, in a context where online learning and AI agents have become commonplace. We conducted an online, asynchronous survey of undergraduate students (n = 443) in introductory to mid-level computing courses at a large Southeastern U.S. university. The survey collected measures of self-efficacy, responses to 13 potential cheating scenarios, the impostor phenomenon (Clance Impostor Phenomenon Scale), and demographic information. We observed a significant negative correlation between self-efficacy and the impostor phenomenon, and a positive correlation between the impostor phenomenon and judging an AI-related scenario as serious cheating. The impostor phenomenon was also correlated with perceived course dedication and prior teaching assistant (TA) experience, whereas self-efficacy was correlated with ethnicity. These findings highlight the need to better understand students' perceptions of AI and to keep the impostor phenomenon present in conversations about academic integrity in higher education scenarios.

Multi-camera streaming perception is increasingly deployed on heterogeneous edge platforms shared with co-resident workloads, yet accelerator placement is often evaluated using isolated single-stream experiments and mean streaming average precision (sAP). Using two end-to-end pipelines on a single GPU–NPU platform, we show that isolated evaluation can mis-rank deployment-time placement. Although the GPU pipeline is preferred in isolation, GPU-localized contention introduces deadline misses that make detections stale and can reverse the preferred placement before full GPU saturation. The NPU pipeline is less accurate than the GPU pipeline on small and medium objects in isolation, but nearly matches it on large objects. The largest absolute sAP losses in our latency and contention experiments occur for large objects. In our four-stream experiments, the preferred placement depends on which path becomes stale, and increasing GPU-side contention shifts the best placement from All-GPU to All-NPU. Under a GPU-saturating vision–language co-tenant, All-NPU achieves $5.2\times$ the worst-stream sAP of All-GPU. Because mean sAP can hide severe single-stream degradation, evaluation should report contention sweeps, deadline-miss rates on both paths, and worst-stream sAP alongside mean sAP.

A linear probe can show that a label is recoverable from a model's hidden states, but not whether that goes beyond what the input already reveals, or whether the model uses it. We audit Toto, an observability forecasting foundation model, on the Benchmark of Observability Metrics (BOOM) across five series-disjoint resplits, comparing linear probes on its frozen residual stream with models that read the raw input window and with Toto's architecture stripped of its trained configuration. Short-vs-medium cadence and metric type are more linearly recoverable from Toto's residuals than from the strongest raw-window model in every resplit (macro-F1 0.766 vs. 0.633 and 0.545 vs. 0.498). Domain is nearly tied, and series cardinality is recovered far better from the raw window. MOMENT-base shows related cadence, metric-type, and domain readouts. Recoverability is not use: exchanging Toto's residuals with those of high-burst donors moves a future-burstiness readout as intended but does not make forecasts consistently burstier than a randomized donor. A BOOM-trained coordination probe has negative zero-shot R^2 on the tested external benchmarks. We report each label against its strongest baseline.

Stochastic order-independent transparency enables efficient and elegant rendering of primitive-based radiance fields like 3D Gaussian Splatting models, but remains impractical due to the inherent visible noise in the output. We propose a principled approach to minimize high-frequency noise, addressing its sources at the representation and image synthesis level. During stochastic rendering, our history-based spatial resampling scheme drastically accelerates image convergence, while temporal importance resampling ensures coherence under camera movement. During training, a color regularizer implicitly reduces the variance along view rays in the 3DGS models. With these properties, our optimized, Vulkan-based renderer effectively mitigates output noise at low and high sample counts, achieving a substantial 13~dB PSNR increase in quality over previous stochastic methods at 1 sample per pixel and quickly converging to sorted 3DGS with an average L1 error of less than $10^{-4}$.

Much of expressive text-to-speech research rests on an untested assumption that written text carries enough information to select an appropriate prosodic style for its delivery. Text-predicted style models improve listener preference, and expressive-appropriateness evaluation presupposes that context constrains style, yet neither measures the assumption itself. This paper tests it as a falsifiable hypothesis against style labels derived from acoustics alone. For each of six speakers in a 1,200-hour conversational corpus, utterances are clustered in the spaces of five speech models, including a prosody-only control, and the cluster of held-out utterances is predicted from twelve text embedding models. Three controls are applied: utterance length is erased from the speech embeddings; accuracy is scored against the majority-class floor of unbalanced clusters rather than uniform chance; and a bag-of-words baseline measures word identity alone. Text predicts the cluster above that floor for all six speakers (+0.111 top-3 accuracy), but bag-of-words achieves three quarters of this. Sentence embeddings add only +0.026, largest for encoders not trained for sentence semantics and reversed by tree-based probes for all others. Acoustic clusters are not compact in text embedding space in any of 360 configurations. The prosody-only space weakens the association for five speakers, but not for the speaker showing it most strongly. Text thus informs these delivery clusters mainly through word choice, whether as a cue to prosody or as a marker of topic and recording situation, and reference-free style selection cannot assume more.

We study which structural results about Aronszajn trees require the Proper Forcing Axiom and which require only MA(aleph1). In particular, we show that MA(aleph1) is consistent with the existence of 2^aleph1 pairwise non-club-isomorphic coherent A-trees.