Papers

Adapting vision-language navigation (VLN) policies to new environments is expensive because every additional route and instruction requires an embodied demonstration. Yet standard observation-to-action training uses only a small fraction of the information already contained in each trajectory. In particular, future observations reveal the instruction-relevant landmarks that the agent will encounter, including what they look like and how they are arranged in 3D. We introduce FINE, a Future-Informed Navigation Encoding framework that extracts this latent supervision from existing demonstrations. FINE equips a VLN backbone with two complementary auxiliary representations. First, explicit landmark tokens follow the ordered landmarks specified by the instruction and are trained to predict future landmark regions in both semantic 2D patch-feature space and viewpoint-dependent 3D geometric feature space. Second, an implicit future token learns to distinguish the landmark state that is actually reached from plausible same-scene counterfactual futures generated by a video world model. On R2R-CE and RxR-CE val-unseen, FINE improves InternVLA-N1 by 2.6 and 4.5 success-rate points, respectively, at full training data. More importantly, as demonstrations become limited, the benefit grows: at a 70% demonstration budget, FINE improves success rate by 6.8 points, recovering roughly one-third of the performance lost by reducing the training demonstrations. Project page is available at https://finevln.github.io/.

Deep Differentiable Logic Gate Networks (LGNs) and Lookup Table Networks (LUTNs) offer a promising approach for very low power inference due to their use of simple binary logic operations instead of arithmetic. In this work, we generalize the logic gates of LGNs to more than two input pins, naturally arriving at networks consisting of $N$-input LUTs. To obtain a differentiable expression for training the $N$-LUT entries, we adopt the Boolean equation of a $2^N$:1 multiplexer (MUX) and optimize its input parameters during training. We investigate the applicability of LGNs and LUTNs to inter-patient ECG arrhythmia classification using the MIT-BIH data set. The proposed models achieve up to 94.41% accuracy and a $j\kappa$ index of 0.683 on a four-class task, showing a competitive performance compared to existing CNN-, SVM- and SNN-based methods. Our LGNs and LUTNs only require an estimated 2.89k to 6.17k FLOPs, including preprocessing and readout, which is three to six orders of magnitude less than state-of-the-art methods. We verified our design, which consists of the preprocessing pipeline and a 6-LUTN classifier, by implementing it on a Xilinx Zynq-7000 ZedBoard. The complete system consumes a dynamic energy of 8.25 $\mu$J/inference, of which only 0.46 nJ is utilized by the LUTN classifier. These results show that both LGNs and LUTNs can be employed as lightweight hardware-based classifiers for inter-patient ECG arrhythmia classification.

We study a non-isothermal model for incompressible magnetoviscoelastic fluids in $\mathbb{R}^N$, $N=2,3$, coupling the incompressible Navier-Stokes equations with the evolution equations for the deformation tensor, temperature, and magnetization. The material coefficients are allowed to depend on the temperature, and no artificial diffusion is imposed on the deformation tensor, leading to a coupled hyperbolic-parabolic system. We establish local well-posedness of strong solutions for sufficiently regular initial data and prove global existence for small perturbations of a constant equilibrium. The global analysis relies on a suitable auxiliary variable combining the velocity and the inverse deformation tensor, which reveals a hidden dissipative structure and allows us to derive uniform higher-order energy estimates for the coupled system.

While Large Language Model (LLM)-based agents demonstrate strong capabilities in long-horizon tasks by interleaving reasoning with external environment interactions, the continuous accumulation of context rapidly creates a critical memory bottleneck. Existing memory compression methods rely on task-specific optimization or external auxiliary models, introducing significant computational overhead. Furthermore, the resulting compressed representations tend to lose structured relationships, leading to information dilution, attention collapse, and degraded decision consistency. To address these limitations, we propose Auto-Regressive State Machine (ARSM), a lightweight training-free framework that enables in-situ reasoning compression through structured state evolution. ARSM introduces two key components: (i) a trajectory abstraction mechanism that reorganizes interaction histories into compact Hypothesis-Action-Result (HAR) micro-chains; (ii) a dynamic state machine that regulates hierarchical memory through atomic operations and a compression-control parameter. These components are unified within an auto-regressive, self-compressive generation space, where each model output jointly performs external action execution and internal state updates. We evaluate ARSM on Webshop, Multi-Objective Multi-Hop QA, and SWE-Bench Lite datasets. Experimental results show that ARSM maintains the task performance while simultaneously reducing token consumption, offering a practical, cost-effective route toward scalable autonomous agents for long-horizon tasks.

We prove two dimension-free estimates in Gaussian space. The first is an optimal Meyer-type inequality for the Gaussian divergence: for $p\geq2$, \[ \|\delta V\|_{L^p(\gamma_d)} \leq \sqrt p\,|\mathbb E V| +Cp\,\|DV\|_{L^p(\gamma_d;\mathrm{HS}_d)}\, . \] Here $\mathrm{HS}_d$ is the space of $d\times d$ matrices with its Hilbert-Schmidt norm. Its main ingredient is a dimension-free weak-type $(1,1)$ bound for the second-order transform $D^2\mathcal N^{-1}$, where $\mathcal N := -\Delta+x\cdot D$ is the Ornstein-Uhlenbeck operator. We also include a direct change-of-variables proof of the weaker bounded derivative divergence inequality in the appendix. The second result is a Gaussian Bernstein-Markov inequality. If $P$ is a polynomial, $p\geq2$, and $m_p := (\mathbb E|g|^p)^{1/p}$ for a standard Gaussian $g$, then \[ \|DP\|_{L^p(\gamma_d;\mathbb R^d)} \leq\frac{2\sqrt{2e\, \mathrm{deg}(P)}}{m_p}\,\|P\|_{L^p(\gamma_d)}\, . \] The factor $2$ is unnecessary when $p$ is even, or $P$ is even or odd. Thus, the constant is of order $\sqrt{\mathrm{deg}(P)/p}$ throughout the range $p\geq2$.

The interpretation of polarized spectral lines requires models that combine statistical equilibrium with polarized radiative transfer accounting for atomic polarization, the Hanle effect, and Zeeman and Paschen-Back splitting. The aim is to provide an open-source Python implementation of the irreducible spherical tensor formalism of Landi Degl`Innocenti and Landolfi (2004), hereafter LL04, for forward synthesis of Stokes profiles with interchangeable atomic descriptions. SolRaT solves the statistical equilibrium and polarized transfer problem for interchangeable multi-term and multi-level atomic descriptions. The multi-term model retains coherences between fine-structure levels of the same term and treats the magnetic Hamiltonian by exact diagonalization, allowing the magnetic splitting to be followed from the linear Zeeman regime to the incomplete and complete Paschen-Back regimes. The widely used multi-level model in turn assumes only the linear Zeeman regime, resulting in a simpler and more developed formalism. The implementation is checked against analytic results from LL04, the Unno-Rachkovsky solution for local thermodynamic equilibrium (LTE) Zeeman transfer, the Hazel2 He I D3 synthesis, and published self-consistent non-LTE scattering benchmarks. The examples show how the intra-term magnetic mixing produces departures from a linear Zeeman multi-level description, and how a J-constrained multi-term approach can be used to isolate this effect on a chosen fine-structure component. Synthesizing the same spectral line with interchangeable multi-term and multi-level descriptions under identical conditions isolates the signatures of intra-term magnetic mixing and inter-J coherences from the linear Zeeman response. Such controlled comparisons are central to developing scattering polarization diagnostics applicable across multiple atomic descriptions.

Learning transferable representations for dynamic text-attributed graphs (DyTAGs) requires models to capture underlying interaction dynamics that persist across domains. However, existing methods tend to overfit to domain-specific structural, temporal, and semantic patterns, limiting edge classification performance under distribution shifts. To expose and address this, we formally establish a leave-one-domain-out (LODO) transfer learning protocol for edge classification on DyTAGs. Under this protocol, we demonstrate that state-of-the-art self-supervised methods for dynamic graph learning perform poorly when transferred to unseen domains. Strikingly, existing methods underperform a structurally and temporally unaware Bag of Events (BoE) model we introduce, which inputs only unordered sequences of node and edge text features. Proceeding from the BoE, we propose Spatio-Temporal Semantic Alignment (STSA), which integrates a spatio-temporal encoder that fuses representations of time deltas and node occurrence frequencies into a unified manifold. STSA is trained with a Contrastive Semantic Forecasting objective, which anchors edge representations to a multi-domain textual latent space initialized by a pretrained language model, providing a robust prior that outperforms BoE and all existing methods we evaluate.

HFT systems are conventionally built as a single-threaded event loop, on the rule that every thread hop adds latency. We test that rule against a measurement study of more than a year of CME market data for the NQ front-month contract, following every packet and matching-engine transaction through the feed's two exchange timestamps, and checking the results against a live production receiver. Packets arrive in near-critical self-exciting clusters that belong to the matching engine's transactions, not to how the exchange packs them. The engine often processes consecutive transactions within a fraction of a microsecond, while the market-data publisher sends at most one packet per publisher period of about 7.5 microseconds, so a burst reaches the receiver as a train of packets one period apart. This yields design principles for HFT systems. First, a receiver that handles each packet within one publisher period never queues on arrivals, however bursty the market; there one thread is best. Second, above that period a queueing tail appears, driven by the timing of transactions, not by packet rate or size, and two threads can be better than one: splitting the servicing chain into two stages on separate threads removes most of the tail at the cost of one hop on the median. Third, only the slowest stage matters, so a split pays only if it shortens it. Fourth, just under the period, where the production receiver runs, the remaining tail comes from multi-message packets and variable service times, and the levers are cost per message and spread of service, not thread count. An analytic framework, a burst-limit throughput identity and an exact reduction of the tandem to a single bottleneck server, supports these results.

We propose sliced Orlicz-Wasserstein (SOW) distance which is a generalization of sliced Wasserstein (SW) distance. SOW replaces the $L^p$ norm in SW with a Luxemburg norm cost induced by an Orlicz function $\phi$. First, we prove that SOW distance is a metric on the space of measures with finite Orlicz norm, and show that it recovers the SW distance when the Orlicz function is $\phi(x)=x^p$. Next, we derive the topological properties of the SOW distance. In particular, we show that convergence under SOW implies weak convergence, and the converse is true under the compact support condition. We then present the theoretical results for estimating the SOW distance. We derive sample complexity for both the distance itself and the powered functional of the distance, and prove their minimax optimality. In addition, we discuss the computational algorithm for approximating the SOW distance by Monte-Carlo estimation and bisection search, as well as the associated approximation error and computational complexity analysis. Our experimental results reveal the superior computational efficiency of SOW compared with Orlicz-Wasserstein (OW) distance. Also, in the experiments, we demonstrate the favorable flexibility of SOW distance over SW in detecting differences between distributions by comparing their performance in two-sample tests and evaluating generative models on image datasets.

Multimodal contrastive learning is a dominant paradigm for learning transferable representations from unlabeled data, but standard objectives primarily capture information that is redundant between modalities. Partial Information Decomposition (PID) shows that task-relevant information in multimodal data decomposes into three components: redundancy shared between modalities, uniqueness specific to each modality, and synergy available only from their joint observation. Recent frameworks extend contrastive learning to capture all three components, yet synergy remains undertrained in practice. We propose SynCo (Synergy Contrastive Learning), a method that directly addresses synergy undertraining through dedicated supervision on an interaction residual. SynCo fits a linear projector to predict the fused representation from independently computed unimodal features, and the resulting interaction residual, which removes the linearly unimodal-predictable component, receives dedicated contrastive supervision at negligible computational cost. On the controlled Trifeature benchmark, SynCo achieves state-of-the-art synergy capture with a $+5.98\%$ gain over the baseline, and on real-world benchmarks from MultiBench, DARai, and MM-IMDb, SynCo consistently outperforms or matches prior methods across diverse modality combinations and task types. The method operates as a plug-in to existing contrastive multimodal frameworks without modifying the underlying fusion architecture and can further improve synergy capture when combined with other methods.

We generalized the seminal work of Law & Eberly, and propose a protocol to generate arbitrary quantum states with a three-level atom as an ancilla in a more efficient manner. This protocol moves two excitations in parallel in each control step, thus reducing the number of evolution steps by half. Full analytical expressions for the control field strengths and timings are derived. The system dynamics is exposed to less chance of error accumulation in the timing and amplitudes of the control fields, and quantitative study in the context of the quantum speed limit is also considered.

Medical ultrasound (US) is inherently degraded by speckle, a granular interference pattern that is often treated as a complex form of noise in image restoration. However, unlike random noise, US speckle originates from coherent scattering within tissue and is therefore highly spatially dependent and deterministic under fixed acquisition conditions, making US speckle suppression fundamentally different from natural image denoising. Because speckle-free US targets are unavailable in practice, self-supervised denoising is necessary. Blind-spot networks (BSN) are the dominant self-supervised paradigm for natural images, but their pixel-wise masking strategy assumes spatially independent noise, an assumption poorly matched to US speckle, which is spatially correlated over multiple pixels rather than pixel-wise independent. To address this mismatch, we propose Mask2Restore, a self-supervised US despeckling framework that reformulates despeckling as contextual inpainting with block-wise masking on single noisy images. Unlike pixel-wise BSN masking, block-wise masking addresses this multi-pixel speckle correlation by removing locally correlated speckle neighborhoods and shifting the reconstruction cues used by the network from adjacent speckle correlations to broader anatomical context. We further introduce cross-resolution context regularization (CRCR), which suppresses residual speckle bias by enforcing consistency across multi-resolution predictions. Experiments on simulated and in vivo carotid US, unseen fine-structure cases, and downstream cardiac segmentation demonstrate improved speckle-detail trade-offs, better preservation of fine anatomical structures, and practical value for subsequent image analysis.

Environmental noise obscures the content cues needed for whisper-to-normal conversion. We propose WhisperVC-AV, which restores acoustic content features while leaving the original WhisperVC conversion module unchanged. Its context-guided restoration module combines temporal acoustic context with synchronized lip features through attention and gated residual correction. Experiments on AISHELL6-Whisper show lower character error rates (CERs) than WhisperVC across three ASR systems, on clean speech and under all six signal-to-noise ratio (SNR) conditions with MUSAN noise. The largest gains occur at 0 dB SNR, where Qwen3-ASR CER falls from 36.29% to 28.88%. WhisperVC-AV also improves predicted speech quality while maintaining speaker similarity. The CER gains extend to unseen background noise without retraining, while visual controls support the use of utterance-specific lip cues. Audio examples are available on our demo page.

We investigate a class of two-dimensional hybrid non-autonomous periodically forced neuron models known as the Meng-Huguet-Rinzel neuron model, with a dynamic threshold and reset mechanism. We focus on the interplay between the continuous subthreshold dynamics and discrete spike-induced resets. The model consists of a linear voltage equation coupled with a threshold variable governed by a nonlinear function of the membrane potential (voltage), and incorporates periodic external forcing in the form of either pulse or rectified sinusoidal currents. We analyze the existence and uniqueness of periodic solutions and prove that the non-hybrid version of the system possesses a unique globally attracting periodic orbit. For the hybrid system, we show that there exists at most one periodic orbit without resets, while numerical simulations indicate the possibility of coexistence of a reset-free periodic orbit and periodic spiking attractors. We also prove that some natural polygons contained in the phase space of the hybrid system form compact positively invariant globally attracting sets of this system. These results provide a rigorous mathematical framework for the analysis of dynamic threshold mechanisms in neuron models and contribute to the theoretical understanding of transitions between subthreshold oscillations and spiking behavior under a time-periodic forcing.

One-step text-guided diffusion editing is efficient but prone to spatially misallocated updates that distort the edited object and alter the background. Existing methods often improve stability by averaging the editing field across timesteps. We instead identify spatial energy misallocation as a distinct and measurable failure mode: across two independent noise draws, the residual field is essentially unrepeatable, making the background field unreliable for direct transport, while its total energy still sets a usable magnitude for the draw at hand. BudEdit turns that magnitude into an explicit budget and reallocates it to edit-relevant regions selected jointly by residual energy and cross-attention, controlling where editing energy is spent rather than averaging over timesteps. The resulting training-free, inversion-free editor spends the budget on transport and reuses it to scale a correction in a lower-noise gated refinement. The budgeted injection field matches its prescribed budget exactly and vanishes on the identified background support, by construction. On PIE-Bench with SD-Turbo, BudEdit outperforms ChordEdit under each method's reported default settings on all 11 evaluated metrics, including a $2.1$\,dB gain in background PSNR, $31$% lower DINO, and $36$% lower LPIPS, while improving all five editing-quality metrics and reporting the lowest runtime in the comparison.

Accurate assessment of Schistosoma japonicum-associated liver fibrosis is essential for disease management and long-term follow-up in endemic regions. Ultrasound provides non-invasive imaging, but complex local echogenic patterns and anatomical structures make fine-grained grading challenging. Existing deep learning methods can predict fibrosis scores, yet directly incorporating acquisition views and regional cues into grading while retaining spatial information for inspection remains an open problem. Here we present VCRE-Fib, a view-conditioned regional evidence framework that integrates anatomical context, local information, and global image assessment for fine-grained ultrasound grading. The framework forms view-conditioned local grading evidence before spatial pooling, uses weak localization to guide its aggregation, and combines it with global predictions. Image-only inference jointly returns a fibrosis score, acquisition view, and candidate abnormal-region map. We developed and evaluated the method on a re-curated cohort of 108,709 ultrasound images from 6,373 patients across 35 centers. On a patient-disjoint test set of 4,107 images from 240 patients across four centers, VCRE-Fib reduced the prespecified composite grading risk by 7.115% relative to SFibAI trained and evaluated on the same data split. Image-level mean absolute error decreased from 0.391 to 0.378, alongside lower patient-max, patient-median, and center-balanced risks. The full model also achieved lower composite grading risk than variants that separately removed view conditioning or weak localization. These results support incorporating anatomical context and regional evidence into ultrasound grading while exposing spatial predictions for inspection alongside severity estimates.

Recommender systems rely heavily on heterogeneous behavioral feedback to infer user preference. Although abundant, these signals are imperfect measurements: the same observed behavior can arise from different underlying states, such as genuine enjoyment, passive consumption, or inattention. The challenge is especially acute in short-form video, where watch-based signals are strongly affected by measurement confounders such as video duration - the same watch time can imply different levels of preference for videos of different lengths, while ratio-based metrics can systematically favor short videos. As a result, optimizing raw engagement can amplify measurement artifacts rather than improving user value. We propose a Factorized Latent Value Model (FLVM) for measuring user preference from heterogeneous behavioral feedback. The model treats observed behaviors as noisy measurements of a low-dimensional, factorized latent value state and uses structured output heads to model heterogeneous feedback signals. A restricted baseline path captures predictable variation from measurement-confounding features such as video duration, user propensity, and session context, while a routed latent path estimates preference-relevant value advantage. The resulting latent value score can be integrated into an existing recommender system as a ranking feature or ranking score. On YouTube Shorts, a major short-form video platform, this model improves offline metrics and lifts a primary viewer enjoyment metric by 2.67% in online A/B tests.

Predicting smooth dynamics and collisions requires modeling continuous evolution and abrupt state changes. We introduce HamiFormer, a dual-expert diffusion field combining whole-window denoising with residual-corrected Hamiltonian propagation. Their mixed-state feedback attenuates the direct contribution of inherited autoregressive error: each mixed state guides subsequent propagation within the jointly refined window. Parallel Local Affine Scan (PLAS) amortizes iterative refinement across rectified-flow steps and evaluates derivatives in parallel across physical time. PLAS's affine symplectic maps achieve lower solver error and runtime than sequential explicit Euler in our evaluation. A Regime Model Tree specializes residuals and routing to balance typical-state accuracy against large tail errors. Our analysis gives conditions for physically consistent refinement and warm-start tracking, and finite-window error bounds under diffusion feedback. In 192-step evaluations, HamiFormer reduces normalized phase-space MSE by 26.3% against PhysiFormer on HamiBalls-1 and 21.4% against DiT on HamiBalls-2, with comparable model capacities. Disjoint-interval comparisons show the lowest late-horizon position and momentum errors among baselines on both datasets. Project page: https://hamiformer.github.io/.

Ultrasound (US) acquisition depends on the operator's ability to interpret anatomy and anticipate how the view will change with probe motion. Many robotic US navigation methods select actions without explicitly predicting these anatomical changes. We propose SonoGraph-WM, an action- and goal-conditioned world model for anticipatory probe navigation. The model represents anatomy as scene graphs (SGs), capturing visible structures, their geometry, and spatial relationships without synthesizing US images. Given a history of SGs and probe poses, a unified Transformer jointly predicts future SGs and poses. A receding-horizon planner recursively imagines candidate trajectories, selects the shortest predicted path reaching a goal graph, and follows it over a short execution horizon before replanning from new observations. To reduce reliance on tracked and anatomically annotated US sequences, we generate aligned SG–pose training data from computed tomography (CT) label maps along surface-constrained probe trajectories. On four held-out CT cases, spatial relation F1 remains above 93% over 20 prediction steps, and closed-loop navigation achieves 77.50% and 75.00% success for the gallbladder and pancreas, respectively, using annotation-derived SGs. In robot–phantom navigation experiments with label-map-derived SGs, the planner reached the target view in 73.7% of trials. These findings support CT-supervised anatomical world modeling for probe planning and highlight the importance of frequent observation updates for reliable navigation. Project Page: https://noseefood.github.io/us-sonograph-wm/

Radiomap blind prediction aims to infer radiomaps from observable representations of the propagation environment and base station configuration without field measurements. In practice, the observable representations are inherently incomplete. Thus, the target radiomap is not fully determined by the inputs when generalizing to unseen configurations or environments. Under incomplete observation, we establish a population-level theory of deterministic radiomap blind prediction that identifies the conditional mean as its optimal target and separates prediction error into reducible predictor approximation and irreducible uncertainty caused by missing physical information. The framework further characterizes the train-test risk gap and the uncertainty reduction enabled by observation enrichment. Building on it, we reveal the dual role of propagation priors: they provide physically grounded guidance, yet their implementable forms may bias the attainable predictor. This motivates RadioDecomp, which treats a prior-guided predictor as a correctable base and learns its remaining predictable discrepancy through residual refinement. To evaluate RadioDecomp across distinct propagation-prior designs, we instantiate it with a feature-guided monolithic base and a LoS-Shadow structured base, yielding RadioFR and RadioLSR, respectively. Across random, cross-configuration, and cross-environment settings, experiments confirm the benefit of propagation-related representations and show that both instantiations improve upon their respective bases. Further controlled studies on base capacity, training-support coverage, and observation coarsening corroborate the proposed analysis.

As AI agents are becoming widely adopted in the financial services industry, careful measurement is essential to understand where they can be reliably deployed and where oversight and professional review remain necessary. Such measurement, however, is constrained by limited access to proprietary or privacy-sensitive data. Existing benchmarks therefore often rely on publicly available data, human- and/or LLM-authored tasks, or simplified settings. We introduce FinancialAuditBench, a benchmark for evaluating agents on financial statement audit tasks, along with a framework for systematically generating synthetic engagements. Our task generation framework leverages differentially private aggregate statistics from historical audits along with audit expertise contributed through over 1,100 hours of benchmark development and review. FinancialAuditBench consists of 90 tasks spanning workpaper completion and review across six synthetic audit engagements, each containing an average of 179 files. Evaluation on eleven frontier models shows that while agents complete substantial portions of staff-level audit tasks well, they sometimes perform inappropriate procedures or produce incorrect documentation. Beyond financial auditing, our framework offers an approach for systematically generating synthetic tasks for model evaluation and training in privacy-sensitive domains.

A subspace design is a collection of subspaces $H_1,\ldots,H_n$ of $\mathbb{F}_q^k$ with the property that no low-dimensional subspace $W$ intersects the collection "too much". Subspace designs and related objects in linear-algebraic pseudorandomness have found a broad range of applications, ranging from list decoding to derandomizing algorithms. We construct explicit strong subspace designs over every finite field. In the extremal case where the co-dimension $t$ of each $H_i$ is equal to the dimension of $W$, for every constant field size our construction attains $n=\Omega(k)$ and matches the probabilistic intersection bound up to a constant factor. All previous constructions required the field size to grow with $t$ (or $k$). Our subspace designs also imply new constructions of rank condensers over arbitrary finite fields. This result is the first to achieve an optimal dependence on $k$ while maintaining both a constant output entropy rate and a constant field size. As an application, we construct lossless rank extractors for linear sources of rank $r$, for all $r < q$, with parameters matching those of Guo, Raj, Shangguan and Zhang (FOCS '26), thereby generalizing their result to prime fields and smaller field sizes. Our construction is based on an algebraic-geometric version of the Parvaresh–Vardy codes (Parvaresh–Vardy FOCS '05, Guruswami ECCC '05), extending the framework underlying the condensers of Guruswami, Umans and Vadhan (JACM '09). We view our construction as a linear-algebraic analysis – tailored to affine sources – of the GUV construction, generalized to functions over algebraic curves. More specifically, inspired by Ta-Shma and Umans (CCC '12) we develop a two-level evaluation scheme, where we first evaluate a function on a curve at extension-field points, and then evaluate a corresponding affine-linear polynomial to obtain outputs over the base field.

A trained neural network can be represented by a parameter vector in high dimensions. Learning distributions over these vectors enables the generation of new models across various tasks and architectures. A central challenge is permutation symmetry: permuting hidden neurons can produce distant parameter vectors representing the same function. This introduces variations that a generative model must account for when learning from trained networks. Existing methods typically address this using networks derived from a common base model or costly approximate neuron alignment. We instead parameterize a flow-matching velocity field with a permutation-equivariant Graph Meta Network, enabling direct learning from independently trained networks without alignment. Extensive experiments show that our method closely reproduces the joint statistics of accuracy, functional similarity, and weight similarity of independently trained collections, providing evidence of generation beyond checkpoint memorization. A single conditional model also generates task-specific networks on heterogeneous architectures and generalizes to unseen hidden-width configurations. On a tabular domain-shift task, intermediate conditioning produces individual networks with performance comparable to logit ensembles across both domains. Taken together, our results show how permutation equivariance enables learning from diverse collections of independently trained networks without permutation alignment.

Over-the-air computation has emerged as a scalable and efficient solution for deploying federated learning algorithms in wireless networks by exploiting waveform superposition for simultaneous model aggregation. Most existing work struggles with heterogeneous fading channels. These approaches either enforce unbiased updates from all devices or allow partial device contributions, requiring careful tuning of the convergence bound to mitigate bias under specific fading models. However, the former significantly amplifies receiver noise due to the weakest channel, whereas the latter is sensitive to fading model mismatch and converges only to a biased objective. To tackle these challenges, we propose FedOAG, which employs algorithmic components to automatically satisfy energy constraints via gradient normalization and evenly mix devices' updates through implicit gossiping. Importantly, FedOAG does not require transmission from all devices, nor does it rely on a specific fading model or knowledge of time-varying statistical channel distributions. We show that FedOAG converges to a stationary point of an unbiased non-convex objective at the best possible rate $O(1/\sqrt{T})$ for any stochastic first-order method. We corroborate our analysis with numerical experiments over dynamic wireless conditions on real-world datasets.

Unified multimodal models combine understanding, generation, and editing within a single network, offering a promising foundation for versatile multimodal applications. However, growing multimodal contexts make KV cache storage and access increasingly costly. Existing KV cache compression methods are typically tailored to specific tasks and single-modality caches, while overlooking changes in cache importance across tasks and timesteps. However, in unified multimodal models, each task involves multiple KV cache types, and both their composition and dynamics differ across tasks. As a result, a single compression policy overlooks task- and type-specific requirements, leading to the loss of critical information and degraded quality across tasks. Based on these findings, we propose UniCache, a training-free framework for task- and type-aware KV cache compression. UniCache identifies the cache segments activated by each task and assigns suitable compression policies through offline calibration. It coordinates their parallel execution under a shared storage budget through attention-guided allocation and task-aware temporal scheduling. Experiments show that UniCache achieves $5\times$ KV cache compression for understanding and editing and $2.5\times$ for generation with negligible quality loss, while increasing throughput by up to $1.78\times$ in long-context settings, significantly improving the practicality of scaling unified multimodal models to longer context.

The authors consider several variants of the notion of nilpotency in pre-Lie algebras and investigate the properties of nilpotent ideals.

We study stable solutions of the Allen-Cahn equation with homogeneous Neumann boundary condition on Riemannian manifolds with boundary in the regime where the transition layers meet the boundary orthogonally. Under an a priori bound for the enhanced second fundamental form in the transition region, we establish uniform second-order Hölder estimates and quantitative mean curvature decay for the nodal hypersurfaces up to the boundary in ambient dimensions at most 10. The proof adapts the technique pioneered by Wang-Wei, adding a further correction determined by the geometry of the ambient boundary. Its leading contribution to mean curvature cancels on the nodal set but survives on non-zero level sets in the transition region, whose mean curvature need not decay uniformly. This phenomenon has no interior analogue and is exhibited by stable solutions constructed by Kowalczyk.

Flow critics learn return distributions by transporting Gaussian noise to Bellman endpoints via continuous velocity fields. While velocity bootstrapping stabilizes training by querying a successor teacher, existing methods face a structural dilemma: on straight paths, no residual-free same-time affine mapping can preserve Gaussian initial noise while maintaining an unbiased target. To overcome this limitation, we introduce Retimed Bellman Flows (ReBF). ReBF queries the teacher critic at a dynamically shifted earlier flow time, aligning intermediate student and teacher trajectories. By combining this retimed clock with fresh, decoupled noise generation, ReBF constructs a provably conditionally unbiased velocity target that preserves the Bellman fixed point and contracts under Wasserstein distances. Empirically, ReBF reduces $W_1$ distance to ground-truth return distributions by up to $7.7\times$ on synthetic MRPs and outperforms existing flow critics across 38 challenging OGBench and D4RL offline reinforcement learning tasks.

Several works have explored multi-agent reinforcement learning (MARL) in LLM collaboration. However, constructing reliable rewards is difficult in practice, as complete and accurate metrics are often unavailable and hard to aggregate. Preference learning provides an alternative by learning from comparative human or AI feedback. Yet, its extension to multi-agent systems remains underexplored. To address this gap, we formulate preference-based multi-agent systems (MAS) from decentralized and centralized collaboration perspectives. We also introduce a general multi-agent preference learning framework (MAPL) to solve these problems. MAPL allows iterative updates by comparing the current solution with decentralized or centralized solutions generated by various agents. We instantiate MAPL using MARL from human feedback (MARLHF) with a learned reward model and multi-agent direct preference optimization (MADPO). Experiments on collaborative writing, coding, tool use, and travel planning show that MAPL can improve collaboration quality and efficiency while approaching the performance of MARL with fixed, well-defined rewards. Within MAPL, MARLHF generally outperforms MADPO on most tasks but remains sensitive to data coverage, agent and comparator models, and the underlying MARL algorithms.

This work presents a fabrication process to realize sub-10 nm-thick, high-aspect-ratio Si fins using high-resistivity float-zone single-crystal Si wafers. Following anisotropic wet etching, a digital etching process based on repeated rapid thermal annealing (RTA) oxidation and vapor-phase hydrofluoric acid (VHF) etching of the resulting oxide was developed, enabling thinning of Si fins into the sub-10 nm regime. Cross-sectional scanning transmission electron microscopy confirms fin thicknesses as thin as 6 nm, although with a tapered profile resulting from the initial anisotropic wet etch. Tapering of the fin was reduced by using an N{_2}/O{_2} (5:1) ambient together with a reduced ramp rate during the temperature ramping for the RTA oxidation process. A thickness variation of 3 nm along a 1.2 {\mu}m tall fin was achieved on a fin with 12 nm near the top and 15 nm near the base. The developed process provides a pathway for fabricating Josephson junctions, setting a foundation toward reproducible and scalable superconducting qubit fabrication with single-crystal Si dielectric barriers using readily available Si processing technology.