Optics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Thursday 10 September
We propose an all-reflective two-mirror projection system for extreme ultraviolet (EUV) lithography operating at exposure wavelengths of $13.5$~nm (Mo/Si) and $11.2$~nm (Ru/Be), delivering a fourfold ($4\times$) demagnification of the periodic mask pattern at a numerical aperture approaching unity ($\mathrm{NA}_{\max} \approx 0.993$). In contrast to conventional EUV projection objectives that incorporate 6–10 aspheric mirrors with an overall optical throughput of less than $15\%$, the proposed design redirects each accepted discrete spatial diffraction order scattered by the mask onto the wafer via a dedicated pair of planar mirror facets. The number of reflections is strictly fixed at two for all accepted orders, retaining $50$–$60\%$ of the power leaving the mask in each accepted order. We derive a spatial geometry providing rigorous optical path length equalization across all diffraction orders, thereby removing order-dependent propagation phase shifts. Individually optimized 30-bilayer Bragg multilayer coatings are designed for each facet using the transfer matrix method combined with global evolutionary optimization algorithms. The architecture is generalized to a three-dimensional vector formulation with a two-dimensionally periodic mask. Utilizing inverse lithography technology, Fourier parameterization, and a differentiable electromagnetic modal waveguide solver, we solve the synthesis problem for binary absorber masks (La absorber on a Ru/Be/Sr multilayer mirror). We demonstrate simulated aerial images of sub-10-nm features on the wafer (isolated peaks with a full width at half maximum (FWHM) of approximately $5.4$~nm and line pairs with a critical dimension of $6$~nm) and find that the two peaks remain resolved for the tested wafer defocus values from $0$ to $5$~nm along the $z$-axis.
Signal Processing
Fold First, Detect Directly: Communication Symbol Detection Without Unfolding for Low-Bitrate Modulo-ADCs
Modulo-folding analog-to-digital converters (MF-ADCs) enable low-dynamic-range quantizers to sample high-amplitude signals without clipping. However, downstream processing traditionally relies on waveform unfolding algorithms, which require high oversampling rates and are highly sensitive to noise. For communication receivers, where the goal is symbol detection rather than signal reconstruction, unfolding is a redundant intermediate step. In this paper, we propose an unfolding-free maximum-likelihood symbol-detection framework for oversampled MF-ADCs in the presence of joint channel and quantization noise. By leveraging modulo wrap cancellation, we derive an exact, single-term Mahalanobis-distance metric that operates directly on folded observations. To handle long sequences, we introduce a parallelized block search algorithm that reduces computational complexity to scale linearly with sequence length. Simulations show our detector significantly outperforms existing unfolding baselines and approaches unclipped conventional ADC performance.
Read the full abstract
Operator Algebras
On AF- and type I-ideals in certain crossed product C$^\ast$-algebras
We study locally finite-dimensional ideals in crossed products of totally disconnected spaces by free actions of the integers and in uniform Roe algebras of exact discrete groups. In the first case, we present a dynamical description of the largest locally finite-dimensional ideal, which turns out to coincide with the intersection of all maximal ideals. In the latter case, we provide a coarse geometric characterization of the locally finite-dimensional compact ideals. Moreover, we show that for crossed products of totally disconnected spaces by free actions of exact groups, the largest type I-ideal is locally finite-dimensional. In the case of uniform Roe algebras, we provide coarse geometric conditions for compact ideals guaranteeing that the ideal is type I and admits an embedding of a UHF-algebra, respectively.
Read the full abstract
Information Theory
Geometric Analysis of Doppler-Based Navigation with Low Earth Orbit Satellites
The increasing vulnerability of Global Navigation Satellite Systems has motivated renewed interest in Doppler-based navigation using low Earth orbit satellites, which can determine position, velocity, clock bias, and clock drift from carrier Doppler measurements. However, the geometric dilution of precision (GDOP) in this eight-state problem behaves fundamentally differently from the GDOP in pseudorange-based navigation. In particular, volume-based satellite selection, effective for pseudorange-based GDOP minimization, has been empirically found to perform poorly for Doppler GDOP. This paper establishes a geometric foundation for Doppler GDOP characterization. A closed-form geometric parameterization of the Doppler measurement Jacobian in terms of elevation, azimuth, inclination, and altitude ratio is derived. It is shown that the clock bias sensitivity depends on elevation, altitude, and the angle between the satellite velocity vector and the line of sight. This sensitivity produces a correlation with the clock drift that no geometric arrangement can remove. A Schur complement decomposition of the eight-state information matrix is performed, yielding an exact GDOP inflation formula, governed by a collinearity coefficient that measures the alignment between the clock bias column and a seven-state subspace. It is proven that the geometric coupling that renders clock bias observable is the same coupling that inflates estimation covariance, and that satellite altitude diversity reduces the collinearity by separating satellites at equal elevation onto distinct sensitivity bands. An example employing a two-shell satellite configuration illustrates the analysis.
Read the full abstract
Machine Learning (Statistics)
A Hilbert-Valued Functional Decomposition Framework for Explaining Time-Dependent Outputs
Feature-based explanations quantify features' influence on model predictions, but are primarily designed for scalar outputs. In many applications, however, outputs are functional or multivariate, such as time-dependent trajectories in demand forecasting. Consequently, existing approaches typically explain each output location independently, ignoring dependencies across the output components. We address this limitation by developing a unified framework for feature-based explanations of time-dependent outputs. Specifically, we generalize functional decomposition to Hilbert-valued prediction functions and extend an existing feature-based explanation framework to this setting. Our framework introduces kernel-based output representations that enable time-dependency-aware explanations at multiple levels of temporal granularity, including time-specific, time-resolved, and time-aggregated, while providing a unified view in which existing methods arise as special cases. We validate our framework on synthetic and real-world data, including intraday financial market volatility prediction and energy demand forecasting.
Read the full abstract
Artificial Intelligence
Memory Compression for High-Fanout Agent Sandboxes
High-fanout agent workloads create a growing memory bottleneck because a single task may spawn many concurrent sandbox sessions. Yet these sandboxes are far from independent: they originate from a shared template and execute related trajectories, exposing substantial template-relative and cross-sandbox memory redundancy. Conventional memory compression is poorly matched to this setting in three fundamental dimensions: how to compress, because they fail to exploit similarity across non-identical sandbox pages; what to compress, because they control page-fault overhead through conservative page selection; and when to compress, because compression is either triggered by memory pressure or performed without awareness of agent execution phases. We present AgentZip, the first memory compression system designed specifically for AI-agent sandboxes. AgentZip introduces compression mechanisms that exploit both the template-relative and cross-sandbox redundancy. It broadens the compression scope to any page with a profitable representation and shifts overhead control from compression-time page selection to restore-time prefetching. It further aligns expensive compression with LLM waiting periods to avoid interfering with foreground tool execution. Across LLM training and inference workloads, AgentZip reduces sandbox-owned memory by up to 8.7x, compared with 2.1x for the Linux configuration. Restore prefetching and agent-execution-aware scheduling reduce the slowdown of aggressive compression from as high as 3.1x to 1.40x while retaining nearly all of its memory-saving benefit.
Read the full abstract
Probability
Discrete Hyperbolic Secant Distributions
We introduce a family of discrete hyperbolic secant distributions on $\mathbb{Z}$, whose normalizing constants arise from series values calculated by Ramanujan and are expressed in terms of Gauß' constant $G=\varpi/\pi$, where $\varpi=\Gamma^2(1/4)/(2\sqrt{2\pi})$ is the lemniscate constant. Using the elliptic lambda-star function $\lambda^*$, we construct scaled versions of these distributions parametrized by $\sqrt{r}$ and $1/\sqrt{r}$ for $r \in \mathbb{N}$. For the first such distribution, we compute the moments up to the eighth degree in closed form via Poisson summation, exploiting that the hyperbolic secant is a fixed point of the $-2\pi i$-Fourier transform. As a byproduct, we obtain closed-form values for series of odd powers of the hyperbolic secant up to the ninth degree, e.\,g.\ $\sum_{k \in \mathbb{Z}} \operatorname{sech}^3(\pi k) = (G^3+G)/\sqrt{2}$, and reinterpret several classical Ramanujan series probabilistically. Finally, we apply our results to evaluate a contour integral, on the critical line, of the product of Dirichlet's beta, gamma, and Riemann zeta functions.
Read the full abstract
Optimization and Control
pqSEDMD: Subspace Methods for Extended Dynamic Mode Decomposition Identification
The dynamic mode decomposition (DMD), along with its variant for nonlinear systems, the extended DMD (EDMD), are powerful tools for the extraction of meaningful spatio-temporal characteristics of (non)linear dynamical systems from measurement data. Despite some efforts to handle the identification task when dealing with real-world data, the decomposition based methods face a critical challenge: real-world data has two inherent sources of uncertainty, the process and measurement noise. Subspace identification methods, are robust tools able to provide accurate state-space models for multi-variable linear systems directly from input-output data. Combining these two methods, we introduce the p-q quasi-norm Subspace Extended Dynamic Mode Decomposition (pqSEDMD). An algorithm that uses our previous improvements to the EDMD by the use of a p-q-quasi-norm reduction on an orthogonal polynomial basis, the pqEDMD algorithm, along with subspace identification methods. The result is a robust approximation of nonlinear systems in a linear function space, combining the strengths of the two methodologies. Throughout the paper we will use the Duffing oscillator as a benchmark problem to show the effectiveness of the algorithm and illustrate many important aspects related to the development.
Read the full abstract
Artificial Intelligence
Off-Target Effects of Response-Style Alignment in a Korean 27B Language Model
We post-train Qwen3.8-27B for Korean response style – verbosity, list and markdown usage, discourse structure and register – and measure two behaviours the objective never targets: abstention on ambiguous social questions in KoBBQ, where the benchmark-correct answer is UNKNOWN, and unprompted disclosure in securities guidance. Both move, and the changes are expressed primarily through the model's emission policy: how often it answers and how much it says. Matched target-form controls show that answer propensity depends on the training target, not the prompt set or recipe alone. Holding prompts, recipe, data volume and serving fixed and changing only the target text, three style seeds give positive answer-rate point estimates (mean +0.82 pp) and three neutral seeds negative ones (mean -1.53 pp); the observed seed ranges do not overlap and the means differ by 2.34 pp. A length-matched arm lies between them, and a fourth arm that stays short while preserving hedging is unstable across seeds, so which feature of the form is responsible is unresolved. For absolute stereotyped exposure the decomposition into an answer-propensity term and a conditional-composition term is an algebraic identity, not a finding; its empirical content is where the movement went. Across the trained checkpoints the changes are dominated by answer propensity while the composition term stays small, and because that term is evaluated on treatment-dependent answered subsets we do not read it as evidence about latent preference. Two measurement results follow. A between-arm contrast in conditional stereotyped share does not identify a change in conditional content preference when answer status is treatment-dependent. And agreement between two rule detectors for the same construct runs from 0.44 to 0.99 depending on which checkpoint produced the text – observable without any reference labels.
Read the full abstract
Probability
Some remarks on Talagrand's convolution conjecture
We present self-contained martingale proofs of Talagrand's convolution conjecture in the Gaussian and Boolean settings with explicit dimension-free constants. The argument is based on the analysis of the stopped density martingale and a weighted Itô isometry estimate. The Gaussian case was settled by Eldan and Lee, and Lehec; the Boolean case was proved up to a $\log\log$ factor by Chen and recently settled by Lu, Guo and Fang.
Read the full abstract
Information Theory
Occupancy-Domain Over-the-Air Computation
Over-the-air computation (AirComp) aggregates distributed data through the wireless multiple-access channel, but coherent implementations require channel state information (CSI), phase alignment, and power control, whereas non-coherent energy methods remain affected by fading. Signal superposition at the receive antenna is linear but requires coherence, and energy superposition is linear only in expectation over fading. We introduce occupancy-domain computation (ODC), whose observable is neither a received amplitude nor an energy: the sum is carried by the silence of the shared resources. With exponential Bernoulli activation, the individual silence probabilities multiply, and the server recovers the sum from the idle fraction using binary activity decisions alone, so that once an activation is detected the amplitude that produced it does not enter the estimate. We characterize the maximum-likelihood estimator, optimal load, and a scale-integrated Fisher-information bound for non-adaptive operation over unknown dynamic ranges. We then introduce balanced occupancy computation (BOC), where each device forms a data-dependent quota of random burst placements. This removes the random placement-count fluctuation of Bernoulli activation; under ideal detection, the leading-order asymptotic root-mean-square error of BOC is no larger than that of Bernoulli ODC at any load and approaches $1/\sqrt{2M}$, where $M$ is the number of resource elements, as the number of devices becomes small relative to $M$. We further analyze unknown-scale operation, finite-frame deviations, and heterogeneous detection misses. Simulations validate the theory and compare ODC/BOC with affine non-coherent energy aggregation and REED.
Read the full abstract
Hardware Architecture
Bio-inspired Learning and Decision-Making with Probabilistic In-Memory Computing Hardware: Part 2
This report extends our previous work (Part 1), which introduced an energy-based model for learning and decision-making under uncertainty. The model leverages stochastic Langevin dynamics to continuously evolve approximate probability distributions over neuron states and model weights. However, as noted in Part 1 and confirmed through GPU-based implementations, large-scale probabilistic energy-based models of this nature face significant scalability challenges due to excessive execution latency. This latency stems from a fundamental mismatch: massively parallel models with low arithmetic intensity (such as energy-based models) are being executed on processor architectures like GPUs that rely on high-bandwidth memory (HBM) interfaces. The HBM imposes brutally sequential execution constraints on inherently parallelizable models, creating the false impression that such models are unscalable. In reality, it is the GPU architecture itself, with its dependence on HBM interfaces, that is not a scalable processor architecture for this class of AI model. In this report, we demonstrate using a detailed transaction-level model (TLM) of a probabilistic analogue in-memory computing (AIMC) processor that the same energy-based model can execute well over 1000x faster than data-center-grade hardware by eliminating the HBM interface and performing computation directly within on-chip memory.
Read the full abstract
Algebraic Geometry
Openness of local properties in flat families of superschemes
Recently, Jankowski introduced the Serre criteria for superschemes to define the super versions of normality and reducedness (called generic fermionic regularity and reducedness). We show that for a flat, proper morphism $f:X\rightarrow Y$, where $X$ is a superscheme and $Y$ is a Dedekind scheme of finite type over a field of characteristic $0$, the set of $y\in Y$ for which the fibers $X_y$ are normal, GFRR, regular, and Cohen-Macaulay is open in $Y$. In particular, if one has a degeneration over $\mathbb{A}_{\mathbb{C}}^{1|0}$, this shows that these properties can be lifted from the special fiber to the generic one.
Read the full abstract
Artificial Intelligence
Generating a Consistent Enterprise: Synthesis and Reference-Free Evaluation of Multi-System Business Data
Synthetic relational data is normally produced by a model trained on a real dataset, and its quality is measured as the distance to that dataset. This paper describes a generator that has no real dataset at either end. Given an industry, a company size, a business model, a set of business applications, and a random seed, it produces a complete fictional enterprise: a workforce, a customer base, sales deals, support tickets, recorded calls, chat messages, and documents, all consistent with one another. One entity graph is projected into the native formats of 66 business products, so the same customer appears in the CRM, the support desk, and the call system under one identity. Because no real counterpart exists, realism is built in from cited reference statistics and verified by reference-free measurement: a five-axis scorecard of 28 statistical checks, an adversarial detector that hunts for the marks of synthetic generation, and a set of soundness checks that include a classifier test against an independently shuffled copy of the data. Because these instruments existed before the generator was tuned, progress is measured under a fixed yardstick: over 23 generated companies, mean realism climbed from 60.3 to 99.1, the weakest company from 41.1 to 94.9, and the detector, which initially flagged 55.2% of all records, now flags none. The scores hold on a seed never used during development. A second generator builds relational databases from a list of business questions. It forces qualifying rows for each answerable question, adds controlled near misses, and computes exact labels from the finished tables. The generator runs as a hosted service at https://console.era.eon.io. A company built there to a specification is served through its simulators over MCP and REST, and the simulators are also published as container images for offline use
Read the full abstract
Computational Complexity
Max Independent Set Remains NP-hard when Excluding a Planar Induced Minor (and More)
We show that Max Independent Set (MIS) remains NP-hard in graphs excluding the $5 \times 5$ grid as an induced minor. This negatively answers a question of Dallard–Milanič–Štorgel that has attracted sustained attention, and refutes a conjecture by Gartland and Lokshtanov. Unless $N$-variable 3-SAT can be solved in time $2^{\mathcal O(N^{1-\varepsilon})}$ for some $\varepsilon > 0$, MIS requires $2^{\Omega(n^{1/2-o(1)})}$ time on $n$-vertex graphs in this class. This complements a $2^{\widetilde{O}(n^{2/3})}$-time algorithm on any $H$-induced-minor-free class by Korhonen and Lokshtanov. Exploiting a connection between low twin-width or reduced parameters on one hand and planar-induced-minor-freeness on the other, we further show that 3-Coloring, Feedback Vertex Set and Dominating Set are NP-complete on weakly sparse graphs of reduced component max-leaf (hence twin-width) at most 3, even when a witnessing contraction sequence is given. Classes of bounded reduced component max-leaf (by any constant) and of twin-width at most 3 are known to exclude a fixed planar graph as an induced minor. The 3-Coloring and Feedback Vertex Set results refute more cases of the Gartland–Lokshtanov conjecture, even when the inputs are weakly sparse. The complexity of MIS in weakly sparse graphs excluding a planar induced minor remains open.
Read the full abstract
Analysis of PDEs
A Sharp Mass Threshold for Fractional Choquard Equations with Lower Hardy-Littlewood-Sobolev and \(L^2\)-Critical Terms
For \(\frac12\le s<1\), we study a fractional Choquard equation with prescribed mass and nonlinearities at the lower Hardy-Littlewood-Sobolev and \(L^2\)-critical exponents. The sharp Hardy-Littlewood-Sobolev and Choquard Gagliardo-Nirenberg inequalities determine an explicit critical mass \(a_*\). For \(0<a\le a_*\), we compute the exact infimum of the constrained energy and prove that it is not attained and that no normalized solution exists. For \(a>a_*\), the energy is unbounded from below on the mass sphere, while the Pohozaev set is nonempty.
Read the full abstract
Mathematical Physics
Recent Advances on a Model for the Electromagnetic Dirac Field at Thermal Equilibrium
We review recent progress on a model for the relativistic electron-positron field interacting with a classical electromagnetic field at positive temperature. The ultimate goal is to advance our comprehension of how thermal backgrounds affect polarisation mechanisms of this quantum system.
Read the full abstract
Artificial Intelligence
When Does Text Inform? Benchmarking Information-Theoretic Metrics for Multimodal Time-Series Forecasting
Multimodal forecasting models that combine time series with text annotations promise richer prediction through textual context, but how do we know whether a text annotation meaningfully contributes to the forecasters prediction? This is an information-theoretic question, but to evaluate whether information-theoretic metrics can reliably measure the predictive value an annotation provides, a ground truth benchmark is needed, and none currently exist. We create a synthetic time series signal with annotations in three categories: semantically correct, incorrect, and irrelevant. Because the data generation process is fully controlled, ground-truth information content is known exactly, enabling principled evaluation of six complementary mutual information estimators (KSG, MINE, InfoNCE, CCA, PID and V-information). We show that all six estimators identify correct annotations as most informative, and are able to audit the quality of mixed text corpora, choosing the annotations that result in the best downstream forecasting results without the need for model training. Our benchmark identifies limitations of each estimator, and these are validated on seven real-world datasets, which show how estimator performance differs on weak signals. Finally, we establish practical rules for implementing these metrics for annotation auditing and fusion selection.
Read the full abstract
Artificial Intelligence
Bio-inspired Learning and Decision-Making with Probabilistic In-Memory Computing Hardware: Part 1
Learning and decision-making in animals are often modeled as Bayesian processes, where sensory evidence is integrated with prior beliefs to guide behavior in the face of uncertainty. But what are the inherent neural dynamics that give rise to this ability, and how could they be replicated in computing systems? This abstract discusses a biologically grounded framework in which noisy neural and synaptic dynamics perform inference and learning via stochastic sampling from an internal energy function, capturing uncertainty over latent states and model parameters through neural and synaptic variability, respectively. This enables approaches such as predictive coding networks to account for epistemic uncertainty via Markov chain Monte Carlo sampling. Drawing a parallel between intrinsic noise in biological systems and electrical noise in emerging probabilistic analogue memory technologies, we highlight how analogue in-memory computing hardware naturally emerges as the solution for massively scalable and energy-efficient probabilistic inference.
Read the full abstract
Signal Processing
Sequential Detection-Based Iterative Blind Separation for Single-Channel Co-Frequency Signals
Existing single-channel co-frequency signal blind separation (SCSBS) algorithms struggle to balance separation accuracy, computational complexity, and robustness, while current channel state information (CSI) estimation methods lack precision. To address these limitations, we propose a sequential detection (SD)-based iterative separation (SDIS) algorithm. SDIS incorporates a delayed unscented Kalman filter (DUKF) into an iterative decision feedback framework, jointly enhancing signal separation and CSI estimation. Simulation results show that SDIS outperforms benchmarks in separation accuracy, CSI estimation accuracy, computational efficiency, and robustness. Notably, when the mean bit error rate (MBER) drops below $10^{-4}$, SDIS can tolerate at least $0.8$ dB more noise than the benchmarks.
Read the full abstract
Computer Vision
SAMV-DUSt3R: Instance-Centric 3D Scene Decoupling from Sparse Multi-Views
With the rising demand to decouple objects from 3D scenes, we propose SAMV-DUSt3R, an end-to-end model that injects SAM2 2D masks into MV-DUSt3R reconstruction. A Cross Flow Mask Block uses these masks to steer the network toward the target instance, jointly improving shape accuracy and achieving object-level disentanglement without multi-stage pipelines. To ensure reconstruction stability, a lightweight Spatial RankGNN selects the optimal reference view with a selection accuracy of 73.5%. Extensive experiments demonstrate that our method boosts average reconstruction precision by 11% across various metrics compared to state-of-the-art baselines. These results reveal a strong instance-disentanglement capability and clear benefits for driving, robotics, AR/VR, and heritage digitisation.
Read the full abstract
Statistical Methodology
A unified framework for spatially resolved cortical activation analysis
Cluster-based permutation tests are widely used for analyzing MEG data, even though they are limited to cluster-level inference and do not provide spatially resolved effect estimates. We propose a regression-based framework for modeling brain activity directly on the cortical surface. Spatial effects are represented using Wendland radial basis functions, and model-based gradient boosting is employed for data-driven selection and estimation of localized activation patterns. This yields interpretable, spatially resolved effect estimates while mitigating the need for extensive multiple testing correction. In a simulation study with heterogeneous signal structures, the proposed approach recovers localized effects that are difficult to detect using cluster-based methods. An application to experimental MEG data further illustrates its ability to reveal spatially specific activation patterns. Overall, this unified framework provides a flexible and interpretable approach for modeling cortical surface data within a unified statistical setting.
Read the full abstract
Artificial Intelligence
Predicting Train Delays in Finland Using Machine Learning and Weather Data
Reliable railway operations depend increasingly on real-time environmental intelligence delivered through wireless sensor infrastructures, a capability that 6G networks will substantially enhance through integrated sensing and edge computing. Adverse weather, particularly in Arctic regions with extreme temperatures and heavy precipitation, remains a leading cause of train delays, yet most prediction approaches rely on raw meteorological inputs without exploiting domain-informed feature engineering. This paper investigates machine learning for train delay prediction using the Finland Integrated Train-Weather (FI-TW) dataset, which fuses railway operational records with observations from the Finnish Meteorological Institute's nationwide sensor network of approximately 200 stations communicating over wireless links. We evaluate three feature configurations using XGBoost at Oulu central station (101,146 observations): full weather features, instant weather observations only, and derived weather category scenarios. The category-based approach, employing hierarchical classifications such as Blizzard, Heavy Snow, and Extreme Cold, achieved an R^2 of 0.78, root mean squared error of 8.5 minutes, and mean absolute error of 3.7 minutes, representing an 11% R^2 improvement and 10% error reduction over alternative configurations. These results demonstrate that compact, domain-informed features derived from sensor streams outperform raw meteorological observations, offering bandwidth-efficient representations suitable for edge deployment over current and emerging wireless infrastructures.
Read the full abstract
Number Theory
The irrationality measure of {\pi} is at most 7.101862832357
We introduce two independent numerator exponents into the Zeilberger–Zudilin integral and specialize them to \[ A_1=A_2=\frac{1857}{2785}. \] The resulting integer linear forms in $1$ and $\pi$ prove \[ \mu(\pi)<7.101862832357. \] This lowers the Zeilberger–Zudilin upper bound $7.103205334137\ldots$ by more than $0.001342501780$; the difference between the unrounded bounds is $0.0013425017806509\ldots$, approximately $0.01890\%$. The same parameter point is a strict two-dimensional local minimizer of the explicit auxiliary upper-bound function in its admissible arithmetic chamber. This is a local statement about that function, not a claim that the point is a global optimizer among all constructions.
Read the full abstract
Cryptography and Security
Few-Shot Learning for Network Intrusion Detection: Methods, Datasets, and Performance
Anomaly-based network intrusion detection systems (NIDS) are an important first line of defense. However, training NIDS for new attack types is challenging, because labeled attack data are rarely available. Few-shot learning (FSL) addresses this problem by learning from few samples. However, the approaches and evaluation settings, that have been investigated so far, vary widely. This work systematically reviews FSL approaches for NIDS published from 2022 to 2026. We conduct a systematic literature review with PRISMA 2020-like reporting to search ACM Digital Library, IEEE Xplore, and Scopus. From a set of 1,358 initial records, we retain 21 studies after screening, deduplication, and quality filtering. We classify the applied FSL approaches, datasets, and experimental parameters and compare reported performance. Meta-learning and convolutional neural networks are the most common approaches, with 8 and 10 studies, respectively. Most studies evaluate five or fewer samples per class, although settings vary. CIC-IDS2017 and CSE-CIC-IDS2018 are the most frequently used datasets. Missing parameters and source code limit reproducibility and direct comparison between approaches.
Read the full abstract
Sound
Xiaomi-CocktailASR-1 Technical Report
Recently, large language model (LLM) based ASR models have achieved significant progress, yet they generally lack support for multi-speaker scenarios, where the cocktail party problem remains a critical bottleneck for further advancing ASR. Existing TS-ASR methods, including end-to-end architectures with speaker embeddings and latest LLM-based explorations suffer from degraded single-speaker performance and the inability to reject when the target speaker is absent. In this paper, we propose Xiaomi-CocktailASR-1, an LLM-based end-to-end TS-ASR architecture. By utilizing reference speech as voiceprint prompts, it directly transcribes the target speaker's speech without requiring speech separation. Xiaomi-CocktailASR-1 maintains competitive performance in single-speaker scenarios, comparable to mainstream ASR models. It also features a negative sample rejection capability, outputting empty text when the target speaker is absent from the mixed speech. Additionally, Xiaomi-CocktailASR-1 supports a Chain-of-Thought (CoT) reasoning mode to provide explicit reasoning steps. Extensive experiments on various synthetic and real-world multispeaker benchmarks demonstrate that Xiaomi-CocktailASR-1 achieves state-of-the-art performance, effectively addressing the cocktail party problem through a unified architecture that balances multispeaker and single-speaker recognition accuracy, along with rejection capability.
Read the full abstract
Numerical Analysis
Minimax-Optimal Early Stopping for SGD via the Discrepancy Principle
We study early stopping for stochastic gradient descent with a constant step size applied to noisy linear inverse problems with polynomially decaying spectrum. The iteration is stopped by the discrepancy principle, that is, as soon as its residual reaches the prescribed noise level. For Hölder source conditions below a saturation level determined by the polynomial decay, we prove that the mean-square reconstruction error at the stopping index is bounded by the classical rate, up to an explicit factor involving the step size and the incoherence of the design, and that the expected number of iterations is of the order of the oracle stopping index. The rate is minimax-optimal for deterministic noise whenever the spectrum contains an eigenvalue at the oracle scale and this factor stays bounded. The step size is restricted by a spectrum-weighted incoherence condition. The analysis allows rank-deficient designs and inconsistent observations.
Read the full abstract
High Energy Physics (Theory)
Graviton vertex operators in the de Donder gauge and BRST descent
We revisit the role of the trace sector of the graviton in covariant string amplitudes by constructing the ghost-number-three graviton vertex operator via BRST descent, imposing only the de Donder gauge condition without requiring tracelessness. The resulting operator is well defined even at zero momentum, while for nonzero momentum the trace-dependent sector of the descent chain forms a trivial BRST descent chain. Evaluating its disk one-point function in the presence of a D$p$-brane, we show that the trace-dependent term is essential for reproducing the correct graviton-brane coupling, canceling the Dirichlet-direction contributions and leaving the expected coupling to the world-volume components.
Read the full abstract
Computer Vision
Order-Aware 2.5D Multiple Instance Learning for Preoperative MRI-Based Perineural Invasion Risk Assessment in Intrahepatic Cholangiocarcinoma
Perineural invasion (PNI) is an adverse histopathologic marker in intrahepatic cholangiocarcinoma (ICC), but it is usually confirmed only after resection. Preoperative T2-weighted MRI may provide noninvasive imaging cues predictive of PNI, although labels are available only at the patient level without slice- or voxel-level annotations. We propose Order-Aware Slab Multiple Instance Learning (OAS-MIL), a weakly supervised framework for patient-level PNI prediction. Each tumor-centered MRI crop is represented as an ordered sequence of overlapping 2.5D slabs formed from contiguous axial slices. A shared encoder extracts slab-level features, which are aggregated by a permutation-invariant set-attention branch and a bidirectional sequence-attention branch. Using five-fold label-stratified cross-validation at the patient level, OAS-MIL achieved a mean AUROC of 0.770, outperforming the evaluated volumetric and MIL baselines. These results suggest that axial order provides a useful inductive bias for weakly supervised PNI prediction from MRI.
Read the full abstract
Robotics
Beyond Noise Steering: Dual-Latent Space Reinforcement Learning for Generative Robot Policy
Pretrained generative robot policies learn expressive action priors from demonstrations. However, existing reinforcement learning methods only steer the noisy space but fail to modulate intermediate action representations during the generation process, resulting in performance degradation and inefficiency. To address this limitation, we propose a novel Dual-Latent Space Reinforcement Learning (DLSRL) framework, which complements initial-noise steering with representation-level control inside the frozen generator. Specifically, our actor network predicts two distinct latent variables: an initial-noise latent variable that steers behavior generation, and an action-representation latent variable for intermediate feature modulation. Moreover, this representation latent variable is mapped to adapter features and ingeniously injected into the hidden states of intermediate action tokens via residual connections. Our dual-control design enables direct adjustment of action representations without updating the base policy. Experiments across generative policy architectures and robotic manipulation tasks show that DLSRL effectively accelerates online robot policy adaptation and achieves competitive performance. Our code is available at \href{https://github.com/xianchaoxiu/DLSRL}{https://github.com/xianchaoxiu/DLSRL}.