Peer Revenue Past citations are no guarantee of future returns

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Monday 14 September

Computers and Society

Do job seekers value procedure in AI hiring only for error correction? Evidence from a conjoint experiment

Employers increasingly delegate initial screening to automated systems, which in many cases reject an application before any human reads it. Acceptance of such systems plausibly depends both on how well they perform and on the procedure that produces the decision. Prior studies rarely vary procedure and performance independently, leaving it unclear whether applicants value procedure for its own sake or for the errors it corrects. In a preregistered paired-profile conjoint experiment, 1,919 United States job seekers made eight choices between systems with independently randomized levels of decision authority, error rate, explanation, opt-out, appeal, and independent bias audit. The value of the appeal, the opt-out, and the bias audit did not rise as wrongful rejections became more common, each staying within a preregistered equivalence bound. Human involvement carried more weight than any procedural feature, moving stated choice about as much as cutting wrongful rejections from 30% to 10%. These patterns constrain a simple error-correction account and are consistent with applicants valuing procedure partly for its own sake, so that improving a system's performance does not substitute for a right applicants can invoke.
Read the full abstract

Programming Languages

Exo-GPU: Safe, Imperative, User-schedulable Programming for Tensor Cores

Modern GPUs require not only SIMT-style parallelism but also software-managed concurrency between compute and data movement to reach maximum performance. Performance engineers must reason about subdividing work into the hierarchy of computation resources (threads, warps, warpgroups, blocks, clusters), and, in many cases, also must use asynchronous tensor core and memcpy instructions on different levels of the memory hierarchy (registers, tensor core accumulators, shared memory, global memory). Unlike CPUs, where out-of-order execution is managed by hardware and hidden from programmers, GPUs expose explicit instruction reordering to software through these asynchronous instructions. Well-established GPU programming languages generally offer either direct low-level control without safety guarantees (e.g., CUDA C++ inline assembly or intrinsics) or easier-to-analyze, high-level abstractions (e.g., Triton's tile-based model) that hide asynchronous instructions in the compiler backend, which may prevent performance engineers from maximizing performance by tuning critical details. We propose Exo-GPU, an imperative, low-level language that creates minimal abstraction over CUDA. Our key idea is to treat parallelism and synchronization as mere annotations on sequential code rather than as fundamental control flow primitives, enabling verification that these constructs do not alter the program semantics. The benefit is twofold: programmers can reason about code without hidden control flow or mutation, while allowing the Exo-GPU compiler to verify sequential-parallel equivalence–guaranteeing that parallel execution is functionally equivalent to its sequential interpretation. We used Exo-GPU to author GEMM kernels for the H100 GPU, using wgmma, TMA, and split-k. Our kernels achieved over 80% of theoretical peak on large problem sizes, in some cases outperforming the vendor-provided CUBLAS library.
Read the full abstract

Metric Geometry

A doubling Loewner sphere that is not a quasisphere

We construct a metric 2-sphere of finite Hausdorff 2-measure that is doubling, linearly locally connected and Loewner but cannot be mapped to the standard sphere by a quasisymmetric mapping. Our result gives a negative answer to a question of Ntalampekos.

Number Theory

Local Langlands functoriality for Yu's supercuspidals II: Fargues–Scholze's parametrization

This is the second of two papers dedicated to the explicit computation of the Fargues–Scholze correspondence. We compute the Tate cohomology of cuspidal representations arising from Yu's construction. Combining this with modular functoriality in the Local Langlands Correspondence, independence of $\ell$, and the partial characterization of the Local Langlands Correspondence established in the first paper, we compare the Fargues–Scholze and Kaletha parametrizations. Among the consequences, we deduce that the Fargues–Scholze correspondence has finite fibers and is surjective onto inertial L-parameters.
Read the full abstract

Chemical Physics

Site-Selective Functional Group Classification of Auger-Electron Spectra and Core-Electron Binding Energies with Convolutional Neural Networks

X-ray spectroscopy techniques probe chemical states in systems with atom-site specificity; however, the site-selective characterization of local bond environments in novel research materials often relies on a combination of reference spectra, multiple complementary spectroscopic techniques and electronic structure calculations. Auger-electron spectroscopy has long accompanied x-ray photoelectron spectroscopy as a second modality to resolve chemical states with overlapping core-electron binding energies. Here I show that the wealth of information encoded in the Auger spectrum offers the opportunity to train convolutional neural networks for site-selective functional group classification in organic molecules directly from the Auger spectrum lineshape. Furthermore, the inclusion of the core-electron binding energy as an additional modality improves the classification, either by augmenting the input fitted intensities with the binding energy or by conditioning the classification on the binding energy via feature-wise linear modulation layers. The latter approach was previously developed for visual reasoning in image classification, and the present results show that this is the more robust approach for including the binding energy. This work demonstrates the potential for new data-driven characterization capabilities in spectroscopic techniques that are information-rich but difficult to interpret.
Read the full abstract

Graphics

Dichoptic Foveation

Interocular differences in visual perception can induce a variety of effects when fused by the brain. For example, prior works have found that carefully crafted binocular differences in local detail can improve contrast. It has also been found that when the frequency content of two stimuli are slightly different, blur suppression leads to a fused percept that is typically dominated by the sharper image. In this paper, we develop a psychophysical framework to measure the perception of natural image stimuli with interocular frequency differences across the visual field. To this end, we study the effect of dichoptic foveation, which we define as the application of blur to one eye and a simultaneous sharpening filter to the other. Stimuli were viewed in a virtual reality (VR) head-mounted display (HMD) and placed at different retinal eccentricities. Study data were scaled to a perceptual just objectionable difference (JOD) scale, and a 4D model was fit to it; our results suggest that interocular frequency differences can be well described by a simple computational model. We applied the model in a realistic VR scenario with free exploration of 360{\deg} videos to improve a base foveated rendering system by enhancing high frequency information dichoptically.
Read the full abstract

All authors Henry Kam, Colin Groth, Jenna Kang, Pratham Saraf, Qi Sun, Kenneth Chen

Astrophysics of Galaxies

NOCTURNE. II. Extreme radio variability in the heart of early-stage active galactic nuclei

We report the discovery of extreme 37 GHz radio variability in early-stage active galactic nuclei (AGN). Most of these sources belong to the class of narrow-line Seyfert 1 (NLS1) galaxies, which harbour fast-growing, low-mass supermassive black holes that accrete at high Eddington ratios. At 37 GHz, these extraordinary AGN exhibit amplitude variability of three to four orders of magnitude over timescales of a few days. Notably, despite several attempts, we have not detected relativistic jets in these sources, suggesting that this phenomenon may occur very close to the black hole and may represent a new form of AGN variability. One of these sources, which flared at 37 GHz, was followed up with the Karl G. Jansky Very Large Array and Swift XRT/UVOT. Based on these observations, we estimated an e-folding timescale of mere hours, leading to variability brightness temperatures and variability Doppler factors that are extremely difficult to explain by an incoherent emitter. This suggests a possible detection of coherent emission from an AGN, although more observations will be needed to distinguish between alternative scenarios.
Read the full abstract

All authors Emilia Järvelä, Patrizia Romano, Anne Lähteenmäki, Marco Berton

High Energy Physics (Phenomenology)

Topological Integer-Winding Dark Matter with a Dark Higgs Triplet: Stability from $\pi_2(\mathrm{SU}(3)/\mathrm{SO}(3)) = \mathbb{Z}$

We propose a dark matter model in which the symmetry-breaking pattern SU(3)_dark -> SO(3)_dark x Z_2 produces absolutely stable point-like topological defects classified by pi_2(SU(3)/[SO(3) x Z_2]) = Z. The lightest neutral member chi^0 of the resulting dark octet, with mass M(chi^0) = 165.67 GeV, couples to Standard Model fields only through a kinetic-mixing portal (g_f = 10^-3) and gravity. A dark Higgs triplet Delta_dark with vacuum expectation value v_Delta = 17 GeV and Yukawa coupling y_chi = 0.10 generates a tree-level splitting Delta M(chi^+ - chi^0) = 1.70 GeV, opening the decay chi^+ -> chi^0 + phi^+ and ensuring that chi^0 is the unique stable relic. An extended Gell-Mann-Okubo mass operator incorporating the quantum number darkicity T_dark predicts three scale-independent mass ratios, r_21 = M(chi^++)/M(chi^0) = 1.089 +/- 0.005, r_31 = M(Sigma^0_dark)/M(chi^0) = 1.163 +/- 0.006, and r_41 = M(Lambda_dark)/M(chi^0) = 1.260 +/- 0.008, falsifiable at FCC-ee. Eight coupled Boltzmann equations with co-annihilation yield Omega_DM h^2 = 0.120 at M_0 = 208.67 GeV. Self-interaction cross-sections satisfy all astrophysical bounds from dwarf galaxies to the Bullet Cluster. The model predicts a stochastic gravitational-wave background at f_* = 1.65 mHz (LISA, SNR = 8.5) and dark acoustic oscillations at r_DAO = 2.3 +/- 0.4 Mpc (SKA).
Read the full abstract

Machine Learning

Attention Mean Fields Predict Average Representation Dynamics and Reveal Context-Specific Computation

A language model's representation geometry is not predetermined; it evolves as the model runs. A faithful account of that geometry must capture that dynamic process, and so cannot be based solely on model-independent statistics such as co-occurrence. Here we introduce a mean-field analysis of attention. The average attention from one token to another defines a kernel that carries representations layer to layer and can be iterated through the network to model how the geometry is transformed. We condition this average two ways. Conditioned on a whole corpus, the kernel predicts the average-case evolution of representation geometry. Conditioned instead on a single context, it predicts the expected geometry for that context. A head's departure from that prediction, its mean-field deviation, isolates the context-specific computation that the mean field misses. Under the corpus-conditional reading, the kernel yields an open-loop model: from the input embeddings and the frozen weights alone, we can iterate the kernel and the model's own MLPs over token representations, never consulting a measured deviation at any layer. The resulting prediction is highly accurate. In early training the model and its corpus mean field are indistinguishable. Replace every attention head with its mean field, and the substitution leaves the loss on real text unchanged. Around the onset of induction, the two diverge, and the gap widens as representations become contextualized. Under the context-conditional reading, deviation from the mean field is a task-agnostic measure of context-specific computation. The residual decomposes additively into unusual attention routing and contextualization of the transported values. Across controlled induction and few-shot settings, greater deviation tracks greater reliance on in-context information.
Read the full abstract

Number Theory

Local Langlands functoriality for Yu's supercuspidals I: Kaletha's parametrization

This is the first of two papers dedicated to the explicit computation of the Fargues–Scholze correspondence. In this first part, we construct (inspired by, and extending, Kaletha's explicit Local Langlands Correspondence for non-singular supercuspidal representations) semisimple inertial L-parameters for all cuspidal representations arising from Yu's constructions, with arbitrary coefficients. We then establish a characterization of this correpondence in terms of certain functoriality properties. This characterization will be used in the sequel paper in order to compute the Fargues–Scholze correspondence after restriction to inertia.
Read the full abstract

Machine Learning

Bounded Adjustment with Reliability-Guided Embedding for Imbalanced Learning with Noisy Labels

Class-balanced learning and label noise create a coupled failure mode: frequency correction prevents majority classes from dominating the decision rule, but can amplify incorrectly labeled minority examples. We introduce BARGE (Bounded Adjustment with Reliability-Guided Embeddings), a single-stage objective combining a bounded, prior-adjusted density-power score with reliability-guided angular geometry. Its classification score is strictly proper in the adjusted probability space and recovers balanced Bayes ordering under clean supervision and the true class prior. Under label contamination, its finite range bounds classification-risk perturbation at a fixed predictor, while its logit gradient redescends when the model confidently contradicts the supplied label. The adjusted target probability also weights class-equal feature compactness, and a one-sided separation term discourages aligned class directions. BARGE requires neither a noise rate nor a transition matrix, uses one network, and leaves inference unchanged. We evaluate it on CIFAR-10, CIFAR-100, and Tiny ImageNet under long-tail and step imbalance, clean labels, and 20% and 40% random incorrect-label replacement. Across 12 clean settings, BARGE ranks second overall and attains the lowest error in four. Under corruption, it achieves the lowest mean balanced error in all six dataset-corruption settings, reducing the six-setting average from 72.32% for the strongest competitor to 70.00%. It also obtains the highest macro-F1 and macro-AUPRC in every corrupted-label setting. Ablations show that class-equal angular compactness improves on the bounded score alone. These results support bounded predictive influence and reliability-guided geometry as complementary mechanisms for imbalanced learning with uncertain labels.
Read the full abstract

All authors Mushir Akhtar, Akarsh J., M. Tanveer, Mohd. Arshad

Optics

Ultimate Performance of Thin-Film Lithium Niobate Mach-Zehnder Modulators

In thin-film lithium niobate Mach-Zehnder modulators, half-wave voltage and bandwidth are tied to the same design parameters; hence, improving one usually costs the other. Several methods have been pursued in the past to alleviate this tradeoff. Specifically, the bias voltage can be reduced through non-centered placement of optical waveguides with respect to electrodes, slow-wave optical gratings, and high-permittivity cladding layers. Bandwidth can be enhanced through implementing periodic segmented electrodes and exploiting low-permittivity bottom cladding and substrate materials. The objective of this work is to study the augmentative impact of these methods. We pursue a rigorous modeling and optimization approach to achieve this objective, estimate the impact of each augmentation, and evaluate the ultimate performance of TFLN MZMs based on these methods. It is shown that a co-designed structure that incorporates all these enhancing methods can attain a drive voltage of 0.84 V and a modulation bandwidth of almost 220 GHz.
Read the full abstract

All authors Shiva Behzadfar, Fatemeh Karami, Pooja Kulkarni, Sasan Fathpour

Robotics

Geometry vs Structure: Graph-Based Diagnostics for LiDAR Point-Cloud Simulation Fidelity

Digital twins provide a scalable and cost-effective complement to real-world testing for validating autonomous-driving and advanced driver-assistance system (ADAS) sensor pipelines. However, quantifying their fidelity remains challenging, particularly for 3D LiDAR point clouds, where conventional geometric metrics may overlook important structural discrepancies. We present a graph-based framework for evaluating the structural fidelity of simulated LiDAR point clouds against real-world scans. While scan-level metrics such as Chamfer distance capture point-wise geometric similarity, they do not explicitly represent connectivity, topology, or object-level organization. Our framework constructs graphs from real and simulated point clouds, applies Louvain community detection to identify spatially coherent subgraphs, and matches corresponding communities using centroid proximity. For each matched pair, we compute $r_\lambda$, a bounded graph-spectral metric motivated by Weyl's inequality, and compare it with density-aware Chamfer distance (CDC) as a geometric baseline. Controlled perturbation experiments demonstrate that $r_\lambda$ is invariant to rigid transformations and robust to sensor noise while remaining sensitive to structural deformation. We evaluate the framework on 50 paired real and simulated LiDAR scans acquired using a Velodyne VLP-32C sensor and CARLA, respectively. The dataset contains more than 1,000 matched communities across four representative classes: vehicles, vegetation, trees, and building walls. The results show that geometric and structural measures capture complementary aspects of simulation fidelity, supporting graph-spectral analysis as an additional diagnostic layer for validating digital twins in ADAS and autonomous-driving applications.
Read the full abstract

Number Theory

Rational Points near Monofractal Curves and the Strong Oscillation Principle

In their previous work devoted to the distribution of rational points near Brownian motion, the authors conjectured the existence of an oscillation principle governing the asymptotic behavior of the number of rational points with bounded denomi\-nators near the graph of a monofractal curve. In this note, a weaker form of this conjecture is shown to hold for a broad class of deterministic fractal curves. These include the classical Takagi and Weierstrass nowhere differentiable functions, and indeed a prevalent (i.e.~"large") class of functions among those which are Hölder continuous with a given exponent of regularity. This constitutes the first instance of deterministic fractal curves for which a precise count of the rational points under consideration is established.
Read the full abstract

Number Theory

Elementary proofs of congruences modulo 5 for overpartitions with restricted odd differences

In 2015, Bringmann, Dousse, Lovejoy, and Mahlburg defined the function $\overline{t}(n)$ to be the number of overpartitions of weight $n$ where (i) the difference between two successive parts may be odd only if the larger part is overlined and (ii) if the smallest part is odd then it is overlined. In their work, they proved that $\overline{t}(n)$ satisfies an elegant congruence modulo 3. Since then, a number of authors have studied arithmetic properties satisfied by $\overline{t}(n)$. In particular, in 2023, Hanson and Smith utilized the theory of modular forms to prove the following two congruences modulo 5: For all $n\geq 0$, \begin{align*} \overline{t}(80n+40)\equiv \overline{t}(80n+60)\equiv 0 \pmod{5}. \end{align*} Our goal in this work is to provide a truly elementary proof of this pair of congruences.
Read the full abstract

Cryptography and Security

gr-PHYSEC: Real-time Channel-based Key Generation for Physical Layer Secure Wireless Communications

Securing wireless communication against eavesdropping is critical, particularly in dynamic and decentralized environments. We present gr-PHYSEC, a new GNU Radio out-of-tree (OOT) module for real-time physical-layer key generation. Unlike traditional key generation that relies on pre-shared secrets or computational complexity, our approach derives symmetric keys from the wireless channel's inherent randomness. We embed a trained neural network within GNU Radio to extract channel features between trusted parties (Alice and Bob) during probe exchanges. These features are quantized into binary keys, reconciled via Reed-Solomon encoding, and further secured with SHA-512 hashing. The generated keys are then directly used to encrypt data. Real-world experiments at the FAU CAAI connected robotics testbed using ADALM Pluto software-defined radios and NVIDIA Jetson Orin validate the approach with ground robotic platforms. Results demonstrate low key disagreement rates and strong randomness, as verified by the NIST test suite for random and pseudorandom number generators for cryptographic applications. This integration showcases how GNU Radio can support real-time AI-driven security solutions, pushing the boundaries of software-defined secure communication. The source code for this project is available at: https://github.com/C2A2-at-Florida-Atlantic-University/gr-PHYSEC
Read the full abstract

Human-Computer Interaction

When a Story Feels Like Mine: How Personalized Narratives and Humor Shape Older Adults' Empathy toward LLM-Generated Peer Health Stories

Peer stories have been shown to boost self-efficacy in older adults' health behavior change. Despite their effectiveness, peer stories are difficult to deploy in health promotion at scale given the difficulty of matching the diverse health concerns and coping styles of heterogeneous older populations. Large language models (LLMs) have been shown to generate authentic narratives, yet how personalization and narrative affective style, such as humor, jointly shape older adults' responses remains unknown. We developed a theory-driven system that generates first-person peer health narratives varying in personalization and humor through a three-stage LLM pipeline grounded in self-efficacy mechanisms. Thirty-one older adults were invited to participate in a within-subjects lab study. Results showed that personalization increased perceived relatability and relevance of peer stories, especially for older adults with lower humor preference. These findings position individual differences in affective styles as a second dimension in designing personalization for LLM-assisted health communication.
Read the full abstract

All authors Kexin Quan, Precious Olalere, Smit Desai, Jessie Chin

Machine Learning

Certified Uncertainty Propagation in One-Shot Federated Bayesian Models via Posterior Event Transport

Probabilistic certification of Bayesian neural networks lower-bounds the posterior probability that a model satisfies a verifier-defined safety property. In one-shot federated Bayesian learning, however, the deployed model is obtained by aggregating parameters drawn from client-specific posterior distributions, so local certificates do not directly guarantee safety of the aggregated model. This paper develops a deployment-consistent certification framework by propagating local posterior events through the deployment aggregation rule, with an exact geometric characterization for Federated Averaging (FedAvg). Each client constructs disjoint hyper-rectangular regions in parameter space and computes their probability masses. The server forms Cartesian products of these regions, maps them through the deployment rule, and retains a product event only when its aggregation image is verified to satisfy the safety property. Under independent client posteriors, each product-event probability factorizes into local masses, and summing verified disjoint events yields a lower bound on safety probability of the deployed model. For FedAvg with nonnegative aggregation coefficients, the image of a Cartesian product of axis-aligned hyper-rectangles is exactly a weighted hyper-rectangle, introducing no set over-approximation. We distinguish the proposed transported-event certificate from direct certification under posterior distributions induced by FedAvg and Product-of-Gaussians aggregation. Experiments on MNIST and Fashion-MNIST under label-Dirichlet heterogeneity show that the transported FedAvg certificate ranges from 22.51% to 46.89%, while direct global certificates range from 72.05% to 91.39%. Results show that predictive accuracy and certifiable safety do not necessarily follow the same trend, and that global posterior constructions can exhibit distinct certification behavior across architectures.
Read the full abstract

All authors Mahyar Mohammadi, Mohammad Hossein Badiei, Abolfazl Yaghmaei, Hamed Kebriaei

Computation and Language

Register Tokens for Bounded-State Reasoning in Diffusion Language Models

Masked diffusion language models (dLLMs) generate text by iteratively denoising masked tokens with bidirectional attention. Extending reasoning across generation chunks normally requires keeping earlier generated text in context. We ask whether a dLLM can instead continue reasoning after that text is cleared, using only a fixed-size carried state. We implement this state as a small number of register tokens: dedicated fixed-position tokens whose continuous hidden states are trained to carry reasoning progress across generation chunks. We post-train dLLMs to decode a chunk of text, clear it while preserving the register values, and continue decoding from the prompt and carried state. In our main comparisons on LLaDA and Dream, registers outperform discrete-text carry on every benchmark, with gains of up to 8.5 points on math and 19.5 points on code. Registers are especially effective for bounded code generation, where correct programs usually span several chunks. Finally, registers can be further refined with reinforcement learning on long-horizon reasoning tasks.
Read the full abstract

All authors Albert Ge, Chandan Singh, Yufan Zhuang, Xiaodong Liu, Jianfeng Gao, Frederic Sala

Soft Condensed Matter

Soft thermal diodes: grafted polymers provide a highly tunable thermal rectification

We explore the heat-rectification properties of a two-phase fluid confined in a nano-chamber with one wall coated with end-grafted polymers using molecular-dynamics simulations. We find a significant thermal diode effect for a wide range of chamber fillings for both very stiff and fully flexible polymers. A stationary heat flux is imposed on the system by fixing the walls at two different temperatures, computing the heat flow as a function of the filling density. The fluid presents a liquid phase, located close to the cold wall, and, for many fillings, a vapor phase in contact with the hot wall. A vapor-liquid interface is also present and located at different distances from the walls, depending on the fluid filling. We study the system by comparing two operational modes: a direct mode, in which the polymers are grafted on the hot wall and exposed to the vapor phase. In the inverse mode, the wall temperatures are swapped and the fluid rearranges to adapt to the interchanged temperature gradient. We calculate the mean heat flow, number density and temperature profiles for the stationary state in both modes. From them, we computed the density and temperature profiles, heat rectification coefficient, and resistivity profiles, for the two studied extreme cases of polymer bending rigidity. We found that, the nano-chamber presents a significant degree of heat rectification with different characteristics and filling density ranges. We describe the conditions to be met to obtain high thermal rectification as regards fluid filling, polymer properties and fluid-polymer affinity. They can be fine-tuned according to the application or material availability. In the direction parallel to the walls the system is easily scalable towards macroscopic sizes, without affecting the thermal rectification. This makes the soft-diode mechanism very versatile in geometry, size and materials choice.
Read the full abstract

Networking

Fast-Convergent Meta-RL via Gradient-Clustered BS Sampling for Edge Caching

Wireless edge caching networks typically consist of many independent Base Stations (BSs), each facing its own request rate and content popularity profile. Training a Reinforcement Learning (RL) caching agent from scratch at every BS forces each agent to relearn, through slow trial and error, a decision problem that is structurally identical across the network. Meta-reinforcement learning removes this redundancy by learning a shared initialization that adapts to any BS in a few local updates; however, meta-training itself becomes the bottleneck at scale: the meta-gradient must be estimated from a small subset of BSs at each meta-iteration, and sampling this subset uniformly at random yields a high-variance estimate, an issue existing meta-RL caching frameworks leave unaddressed. This paper proposes a meta-reinforcement learning framework for caching across independent, non-overlapping BSs that directly targets this bottleneck. Each BS runs a local Proximal Policy Optimization (PPO) agent, formulated as a Semi-Markov Decision Process (SMDP) over content popularity, size, lifetime, and importance, while a shared meta-policy is learned via a Model-Agnostic Meta-Learning (MAML)-style loop. To scale meta-training and accelerate convergence, we introduce gradient-based clustering, which groups BSs by local gradient similarity and draws from every cluster, in proportion to its size, at each meta-iteration. We prove, via an Analysis of Variance (ANOVA)-style decomposition of gradient variance, that this strategy yields a strictly lower-variance meta-gradient estimator than uniform random sampling under BS heterogeneity.
Read the full abstract

Machine Learning

Autonomous Droplet Navigation via Model-Based Reinforcement Learning

Precise manipulation of liquid droplets underpins lab-on-a-chip platforms for diagnostics, chemical synthesis, and biological assays. Yet autonomous droplet transport through confined geometries of varying complexity remains an open challenge. Droplets exhibit contact-angle hysteresis, deformability, and capillary pinning, which make their response to actuation nonlinear and history dependent, that classical controllers and pre-programmed trajectories cannot cope in multi-turn environments. Here we demonstrate autonomous navigation of a liquid droplet through geometries of increasing complexity on a gravity driven (Labyrinth) platform using model-based reinforcement learning. A thin silicone oil film reduces contact-line pinning while two-axis tilt supplies the gravitational driving force, and an overhead camera tracks the droplet in real time. An offline-trained policy discovers effective tilt strategies from limited physical interaction data, without simulation or analytical droplet models. The system operates under partial observability, as oil-film thickness, instantaneous contact angle, and droplet deformation state remain hidden from the controller. Despite these challenges, the learned policy achieves reliable navigation across straight, right-angle, and curved-arc paths, including outside-corner geometries. We further demonstrate that a policy trained on a simpler geometry transfers to complex ones, succeeding zero-shot on right-angle and staircase paths and reaching full success on a curved arc with a fifth of the training data. The findings suggest promising avenues for enabling droplet based microfluidic systems to serve as intelligent chemical laboratories.
Read the full abstract

2609.16369 · 14 Sep · 43 pages, 15 figures, 3 tables including supplementary material. The source code is available via GitHub at https://github.com/rajneeshanand/DropletRunner. An archived version of all supplementary movies has also been uploaded to Google Drive: https://drive.google.com/drive/folders/1ewK2dxWxfzjd4kk6Df4cqKE3-3xCbOgf?usp=sharing

Robotics

UDAV: Uncertainty-Driven Adaptive VLM Waypoint Planner

Vision-language models (VLMs) can generate routes directly from aerial imagery for off-road navigation, but their predictions provide no indication of reliability. We present UDAV, an Uncertainty-Driven Adaptive VLM Waypoint Planner for UAV-guided UGV navigation. UDAV draws multiple stochastic trajectory predictions, selects their medoid as a self-consistent nominal route, and estimates predictive uncertainty from their spatial dispersion. When the maximum uncertainty across interior waypoints exceeds a threshold, UDAV invokes a reconsideration stage; otherwise, it returns the medoid directly. We evaluate UDAV on 400 held-out trajectory queries from two UAV flights. Stochastic medoid selection reduces the mean average displacement error (ADE) from 147.4 pixels for a deterministic prediction to 115.9 pixels. The complete planner achieves a mean ADE of 110.4 pixels, a 25.1% reduction relative to deterministic planning, while producing valid trajectories for all queries. UDAV also yields the lowest 90th- and 95th-percentile errors among all evaluated configurations, including a higher-budget K=10 consensus baseline. Relative to the K=5 medoid, UDAV reduces these errors from 225.3 and 326.0 pixels to 199.0 and 290.8 pixels, respectively. These results demonstrate that stochastic VLM predictions provide both a stronger nominal route and an actionable uncertainty signal for selectively mitigating large planning errors.
Read the full abstract

Hardware Architecture

FINNAS: FINN-Guided Hardware-Aware NAS and Pruning for FPGA Jet Substructure Classification

FPGAs are well suited to deploying quantised neural networks (QNNs) under strict accuracy, latency, and resource constraints; however, identifying efficient model-accelerator combinations commonly requires extensive manual design-space exploration and repeated hardware synthesis. This paper presents FINNAS, a FINN-guided hardware-aware evolutionary neural architecture search framework. FINNAS jointly searches quantised MLP depth, width, and global precision settings, and ranks candidates using proxy validation accuracy together with FINN-estimated LUT usage and latency under a fully parallel mapping. Selected finalists are fully retrained, subjected to post-search unstructured pruning, and validated using RTL simulation and Vivado out-of-context synthesis. On the CERNBox jet substructure classification task, the searched implementations expose competitive accuracy-resource trade-offs. Compared with a manually optimised dense FINN accelerator, a compact FINNAS design improves accuracy from 73.78% to 74.36%, while reducing LUT usage by \(8.5\times\) and RTL-simulation latency by \(1.77\times\). Unstructured pruning further provides consistent LUT and FF reductions across the fully parallel finalists.
Read the full abstract

All authors Eva Chauffour, Changhong Li, Georgios Floros, Shreejith Shanker

Computation and Language

How Humans and LLMs Read Gender into "Gender-Neutral" Physical Descriptions

When foundation models describe people, recent work in AI fairness, accessibility, and ethics recommends avoiding inferred identity labels (e.g., "she", "his") in favor of seemingly "objective" physical descriptions (e.g., "short hair", "a defined jawline"). Yet whether such descriptive language achieves gender-neutral communication remains an open empirical question. To study this, we introduce GAPA (Gender Associations of Physical Attributes), a dataset of 316 common physical attributes drawn from diverse sources, paired with 14,706 gender-association ratings from 304 US-based annotators. Results show that physical descriptions carry structured and graded gender associations among readers, with more consistent and distinctive associations for women and men than for non-binary identities. Next, we evaluate 16 LLMs across model families, sizes, and post-training variants against human ratings. The models partially recover human associations but exhibit systematic alignment biases, including compressed rating distributions, weaker alignment for associations with men, and asymmetric abstention that disproportionately targets the non-binary category. Finally, we release the best-performing proxy model trained to predict humans' gender associations of descriptive language and demonstrate its utility through a sociolinguistic analysis of character descriptions in LitBank. Together, our findings provide the first empirical evidence that seemingly "objective" physical descriptions can retain systematic gender associations in human interpretation, and uncover systematic patterns of model-human misalignment. This challenges the assumption that replacing explicit gender labels with physical descriptions necessarily yields gender-neutral communication, and highlights downstream challenges in using such descriptions to communicate subjective identity categories in human-AI interaction.
Read the full abstract

Machine Learning (Statistics)

Mini-batch Sampling Strategies for Long-Tailed Image Classification: An Empirical Study on CIFAR-100-LT

Real-world datasets often exhibit long-tailed class distributions, where a few head classes contain a large number of training samples while a large number of tail classes have only a few. The composition of each mini-batch, determined by the sampling strategy, governs which classes contribute to the stochastic gradient estimate, and therefore affects convergence behaviour and generalisation across the whole class spectrum. We provide a systematic theoretical and empirical comparison of four mini-batch sampling strategies for long-tailed image classification: uniform instance sampling, class-balanced sampling, square-root sampling, and progressively balanced sampling. We place all four in a unified bias-variance framework describing their effect on gradient estimation, which exposes the tension between unbiased optimisation of the empirical loss and fair representation of rare classes. We then evaluate them under controlled conditions using ResNet-32 on CIFAR-100-LT at three imbalance ratios (rho = 10, 50, 100), with every strategy sharing the same long-tailed subsets and initialisation within a seed. Progressive sampling improves tail-class accuracy by 25% relative to the uniform baseline at rho = 100 (13.5% versus 10.8%), consistently across all three seeds, while its overall accuracy is not distinguishable from that of uniform sampling given the seed-to-seed variation (40.0% versus 39.7%); the tail-class gain, not the overall gain, is the robust effect. At rho = 100, class-balanced sampling degrades accuracy on every class group, including the tail classes it is designed to help, which we attribute to overfitting caused by extreme oversampling of scarce data; at rho = 50 this failure is confined to head and medium classes. These results indicate that when rebalancing is applied during training matters as much as how much rebalancing is applied.
Read the full abstract

Quantum Physics

Pump-Free Microwave-Optical Bell Pair Generation for Teleportation-Based Quantum Transduction

The coherent conversion between microwave and optical photons, known as quantum transduction, is critical for connecting superconducting processors to optical networks. Existing methods are limited by complications associated with optical pumping. We propose a pump-free microwave-optical Bell-pair source for teleportation-based transduction. Using a spin or atomic system resonantly coupled to optical and microwave cavities, the scheme generates loss-robust heralded Bell pairs. Across color centers, atomic ensembles, and phonon-mediated systems, this assembly achieves kilohertz-range heralding rates with high fidelity.
Read the full abstract

All authors Fangxin Li, Jaesung Heo, Zhaoyou Wang, Benjamin Pingault, Xingyu Gao, Tengyang Ruan, Anjun Chu, David D. Awschalom, Andrew N. Cleland, Andrew P. Higginbotham, Alexander A. High, Liang Jiang

Hardware Architecture

FSNIC: A Low-Latency Flow-Based Intrusion Detection Architecture for FPGA SmartNICs

Modern data centres require high-performance networking alongside effective real-time security. Traditional Intrusion Detection Systems (IDS) commonly rely on general-purpose processors and often struggle to inspect high-speed traffic at line rate without introducing latency or performance bottlenecks. Smart Network Interface Cards (NICs) provide an alternative by enabling computation directly within the network data plane. This work presents a machine learning-based IDS implemented within an FPGA-based SmartNIC pipeline. The system integrates P4-based packet parsing with a LogicNets IDS model implemented in RTL, enabling deterministic, low-latency inference. Compared with traditional stateless packet-level classifiers, the proposed stateful flow-based IDS introduces minimal state by aggregating features across packets, capturing behavioural patterns not observable at the packet level. Experimental results on the UNSW-NB15 dataset show that the flow-based IDS improves detection accuracy from 86.92% to 97.68% compared with stateless packet-level classification. We also evaluate the proposed IDS on CICIDS2017 and compare its real-time hardware performance with prior FPGA-based IDS designs. Through hardware-software co-design, the proposed IDS achieves 6~ns inference latency using only 846 LUTs, with no BRAM or DSP usage, demonstrating a low latency and resource efficient implementation.
Read the full abstract

All authors Nise O'Cuill, Changhong Li, Georgios Floros, Shreejith Shanker

Computational Physics

A Gradient-Reconstruction Lattice Boltzmann Method for Compressible Navier–Stokes–Fourier Equations

Reaching compressible flow has usually forced lattice Boltzmann methods to abandon the compact stencil or the strict locality that make them efficient. We give up neither, solving the compressible Navier–Stokes–Fourier equations with a scheme that transports only the conserved mass, momentum and energy. The viscous stress and heat flux depend on gradients of the velocity and temperature. To capture them, existing compressible schemes go beyond a single lattice of the conserved fields. They enlarge the velocity set, carry the stress and heat flux as extra transported fields, place the energy on a separate grid, or give up exact streaming for an off-lattice advection. We instead recover the gradients from the non-equilibrium part of the distributions already present, inside the collision. No field beyond the conserved state is transported, no neighbour is read, streaming stays exact, and every computation runs in single precision. The method carries five fields where a transported-flux scheme carries fourteen, at a fraction of the memory and $4.9$ times the throughput. Verified against the exact Sod and Becker shock solutions and a supersonic Taylor–Green vortex, it tracks the direct-numerical-simulation reference on the kinetic energy and stays within the reference-solver spread on both dissipation rates.
Read the full abstract