Peer Revenue Now accepting positions

Papers

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

All of arXiv

Sunday 8 November

Machine Learning

Data Manipulation Attacks Against Machine Learning Integrated Cyber-Physical Systems

There is a well-established body of work demonstrating that individual machine learning (ML) algorithms can be manipulated or deceived. However, in many real-world systems, multiple ML algorithms are composed to enable automated decision-making, particularly in Cyber-Physical Systems (CPS). Often such systems use a composition of vision, planning, and other ML models to affect their internal state and ultimate take actions in the physical world. While there is a huge literature on the (in)security of ML algorithms, it is not immediately clear how an attacker might leverage their ability to trick individual learners when attacking the composition as a whole. We develop methods for creating an attack policy (a mapping from state to attack) for a data-manipulating attacker affecting a CPS or other state-machine-based AI compositions. We describe our framework, a basic threat model and corresponding methods for computing an optimal attack policy, as well as planned extensions. The goal of this work-in-progress paper is to stimulate conversation and forge new collaborations regarding the security of compositional learning.
Read the full abstract

Analysis of PDEs

Stability and Optimal Decay of Viscous Contact Waves under Large Perturbations

The stability of viscous contact waves is established for the one-dimensional full compressible Navier–Stokes equations under large integrable perturbations. A unique global positive solution is obtained, which converges uniformly to a fixed Eulerian contact profile at the optimal inverse square-root rate in time. The wave-strength threshold depends only on bounds for the initial density perturbation in the first-order Sobolev norm, the energy of the initial velocity perturbation, the thermal relative entropy and the total conserved excess, together with positive lower bounds for the initial density and temperature. It is independent of higher Sobolev norms and the common far-field velocity. No square-integrability condition is imposed on the initial anti-derivatives. Derivative and pressure decomposition estimates are also established for the evolving solution.
Read the full abstract

Metric Geometry

Bounded Spaces without a Geodesic in the Gromov-Hausdorff Class

We construct two countable complete bounded metric spaces at Gromov-Hausdorff distance $1/4$ with no midpoint and hence no shortest curve between them. The proof uses the local existence of optimal correspondences near a space with finitely many self-correspondences of distortion below a fixed positive threshold, each of which is the graph of a bijection.
Read the full abstract

Machine Learning

A Stability Analysis of AdamW: Unstable Equilibria and Non-Convergence

AdamW is ubiquitous in deep learning, yet its behavior remains poorly understood. We analyze its dynamics through the lens of dynamical systems and show that AdamW admits an implicit fixed-point objective: its fixed points coincide with the stationary points of a constrained and regularized optimization problem. However, not all of these fixed points are stable under AdamW's dynamics, and stability depends sensitively on curvature, weight decay, and momentum parameters. Even in one-dimensional settings, equilibria may be unstable, and trajectories can exhibit persistent oscillations rather than convergence to them. We further extend the analysis to higher dimensions, deriving sufficient conditions for local stability of the continuous-time dynamics, and propose a convergent modification that preserves AdamW's fixed points. These results clarify what optimization problem AdamW is associated with, when its convergence can be expected, and how its dynamics could inspire more reliable optimizers.
Read the full abstract

Machine Learning

Boundary Mass in Feature Space as a Label-Budget Diagnostic for Representations

Low-label transfer is often decided by a practical question: after a representation looks useful, how many labels are still needed to learn the downstream boundary? Transferability and probe-quality scores characterize global alignment, likelihood, margins, and calibration. Boundary-near target mass complements these diagnostics by measuring the ambiguous regions where scarce labels are most costly. We propose effective boundary complexity, a label-budget diagnostic for frozen representations and linear probes. It measures mass in a thin strip around the learned boundary and normalizes it by class balance, so dense ambiguous regions and predicted-class imbalance are both counted. We prove that, under a smooth representation map, density change and tangential surface change cancel at first order, leaving boundary-normal stretch or compression as the governing factor. This yields a cached-feature estimator that fits a probe on one split and counts independent held-out features near its boundary. Experiments support the diagnostic at three levels. Controlled transformations match the predicted law across 74 maps with 0.38% median relative error. On real frozen features, normal-direction changes move both measured boundary mass and label demand, while tangent controls preserve the first-order coefficient. A differentiable surrogate reduces measured complexity by 16-36% at matched top-1 utility. Finally, CIFAR-100 and Tiny-ImageNet experiments support boundary mass as a second-stage diagnostic for near-tied representations. Five-way experiments show strong OvR-interface backbone-ranking agreement (Spearman 0.872), while acquisition experiments show high correlations between initial boundary mass and realized label budgets across three sampling policies (0.779-0.830). Code is available at https://github.com/qqaqq-code/boundary_mass.
Read the full abstract

All authors Siming Zhang, Zhehui Shen, Shijie Chen, Yansen Yu, Hang Yu

Applied Physics

Extreme temperatures characterization of Al0.6Ga0.4N channel HEMT

In this work, we fabricate and characterize ultra-wide bandgap Al0.6Ga0.4N channel HEMTs with Pt/Au gate under a wide temperature range of 20 K to 650 K. The devices retained robust modulation capability throughout the measurement, with excellent on/off ratio of 9.7 * 107 and 6.4 * 107 under 20 K and 650 K. The DC output IV measurement was performed, with Id,max of a device with Lsd = 9 {\mu}m measured to be 240 mA/mm and 120 mA/mm under 20 K and 650 K. In addition, low drain and leakage were observed in multiple devices, and the device robustness was evaluated under room temperature after the extreme temperature measurement with minimal degradation observed in Id,max and Rsheet. Moreover, transfer I-V characteristics of the Al0.6Ga0.4N channel HEMTs have also been studied, where a relatively small Vth shift was observed. TLM measurement and transconductance analysis both indicate the presence of weak temperature dependent scattering mechanisms in the device channel. The subthreshold swing (SS) was calculated and observed to increase parallelly to SSideal under investigated temperature, indicates a robust gate structure. Finally, the on/off ratio and Ig of the 9 {\mu}m Lsd device were benchmarked and compared favorably against other reported results. While the results offer insight into the extreme temperature performance of AlGaN channel HEMT with Al composition higher than 50%, it also establishes AlGaN channel HEMT as a promising candidate for extreme temperature electronic applications.
Read the full abstract

All authors Shuwen Xie, Jacklyn Zhu, Swarnav Mukhopadhyay, Yi-Chen Liu, Shubhra Pasayat, Savannah Eisner, Chirag Gupta

Systems and Control

Ultrawideband Polarization-Rotating Antenna for Enhanced Ground Penetrating Radar Detection

Ground penetrating radar (GPR) represents a powerful and widely adopted tool for nondestructive subsurface detection and imaging. To overcome the impaired detection performance caused by polarization mismatch in conventional single-channel GPR systems, this paper presents an innovative ultrawideband polarization-rotating (UPR) antenna that achieves rotating linear polarizations (LPs) by varying the source frequency for enhanced detection. The developed UPR antenna achieves a continuous polarization rotation from $0^\circ$ to $170^\circ$ across 1.2 to 3.2 GHz, as verified by both full-wave simulations and experimental measurements. The antenna is subsequently integrated into a single-channel stepped-frequency GPR system and evaluated against targets of varying geometries in a sandy environment. Compared with a conventional fixed-LP GPR system, the UPR antenna-enabled GPR system delivers markedly more consistent detection performance across all target types, with particularly pronounced enhancements for polarization-sensitive targets. By enabling ultrawideband frequency-dependent polarization agility, the proposed GPR system enhances subsurface target detection while maintaining the simplicity and low cost of a single-channel measurement architecture.
Read the full abstract

All authors Ming Li, Yuki Gao, Shu-Lin Chen, Hai-Han Sun, Y. Jay Guo

Sound

Music Inpainting With Similarity-Guided Latent Diffusion

We propose a music inpainting technique combining similarity matching and a diffusion-based generative model to replace large segments of audio recordings with suitable alternatives. The proposed hybrid method, Similarity-Guided Diffusion Posterior Sampling (SimDPS), retrieves candidate replacement segments from the wider signal based on contextual similarity and uses them to guide the diffusion process toward coherent reconstructions. Being model-agnostic, the framework is compatible with any pre-trained diffusion model operating in either the waveform or latent domain. Objective evaluations demonstrate that SimDPS outperforms widely adopted RePaint and DPS methods across acoustic features and neural-embedding distance metrics. Subjective evaluations on multi-genre music datasets confirm that SimDPS enhances perceptual plausibility compared to previous diffusion-based methods and demonstrates robustness to poorly matched similarity candidates. These findings pave the way for improved music restoration and context-aware audio editing.
Read the full abstract

Robotics

K-PWM: Control-Oriented Structured Bilinear World Models under Partial Observation

World models learn internal simulators of environment dynamics for prediction, planning, and control, yet learning them from limited, noisy, and partial observations remains challenging. While much progress has focused on pixel-based observations, many physical systems instead provide vector-valued measurements that reveal only part of the underlying state. We introduce K-PWM, a probabilistic world model for prediction and control under partial observation. K-PWM couples Koopman-bilinear latent dynamics with a nonlinear observation decoder and performs probabilistic latent-state inference through the learned model via extended Kalman filtering and smoothing, without a separately trained inference network. Its bilinear structure enables exact propagation of the maintained Gaussian belief through the latent transition and efficient rollouts for uncertainty-aware model predictive path integral (MPPI) control. For training K-PWM, we present a generalized expectation conditional maximization (GECM) procedure with closed-form updates for the generators, noise covariances, and latent prior, complemented by gradient-based decoder learning. A Koopman-consistent initialization selects the latent dimension from data and initializes the action-dependent dynamics. With limited training data, K-PWM achieves lower long-horizon prediction error than the evaluated baselines on partially observed Gymnasium MuJoCo systems, while substantially reducing training time relative to the strongest prediction baseline. On high-dimensional MuJoCo Menagerie robots, it achieves competitive tracking accuracy with low per-step computation times, demonstrating potential for real-time deployment. These results support structured probabilistic dynamics as a data- and compute-efficient approach to world modeling and control under partial observation.
Read the full abstract

All authors Santosh Mohan Rajkumar, Sriram Narayanan, Samuel E. Otto, Debdipta Goswami

Machine Learning

Hydra: Towards Transferable Multi-Task Learning on Temporal Graphs

Real-world evolving networks are naturally modeled as temporal graphs (TGs), where capturing temporal dynamics is essential for predicting future graph properties that support downstream decision-making. Existing temporal graph methods have been developed primarily for single-task prediction, and little is known about their generalization across tasks or transfer to unseen networks. This leaves the challenge of multi-task graph property prediction in TGs largely open. We address this challenge by introducing Hydra, a novel architecture that integrates local connectivity features from temporal GNNs with a spectral learning module that captures global connectivity patterns. This design enables joint learning of local and global information under a multi-task objective. In multi-task classification, Hydra achieves an 8.9% relative gain in AUC over the strongest competitor. In multi-task regression, Hydra achieves competitive results in all three tasks, while obtaining the best results in two tasks with a 8.2% relative gain in MAE compared to the strongest baseline. Moreover, Hydra delivers these gains with a 22% reduction in training time compared to temporal transfer models. These results provide the first systematic evidence that multi-task transferable learning on temporal graphs is effective. By delivering consistent top-ranked performance, Hydra highlights multi-task training on temporal graphs as a promising direction toward adaptable foundation models for temporal graphs.
Read the full abstract

All authors Kiarash Shamsi, Farimah Poursafaei, Tran Gia Bao Ngo, Reihaneh Rabbany, Baris Coskunuzer, Guillaume Rabusseau, Michael M. Bronstein, Shenyang Huang, Cuneyt Gurcan Akcora

Neurons and Cognition

From MNH to the Friedman–Komogortsev Method—which changes improve eye-movement classification?

The Friedman-Komogortsev Method (FKM) is a rule-based classifier of fixations, saccades, and post-saccadic oscillations in 1000-Hz recordings, developed from the modified Nyström-Holmqvist algorithm (MNH) and used in several recent studies. FKM changes MNH in many ways at once: it centres MNH's Savitzky-Golay filter, adds a position-based test for whether an apparent post-saccadic oscillation continues the saccade, adds rejection tests, and changes several thresholds. We asked which of these changes alter agreement with human classification. Both algorithms were scored against twenty 26-s GazeBase reading recordings hand-classified by three trained raters, and each change was reverted in turn. FKM agreed better than MNH as MNH was used: saccade Cohen's kappa rose on all 20 recordings (median +0.034), and saccade boundaries moved from 2-3 ms late to the raters' sample. One change produced this: centring the filter, which removes a three-sample delay. Reinstating the delay returned FKM's saccade kappa to MNH's value, and centring MNH's filter alone made MNH match FKM on saccades and exceed it on post-saccadic events. The position-based test changed how saccades with several velocity peaks were divided, splitting 8% of rater-labelled "wavy" saccades into a saccade and a spurious oscillation against 39% for delay-corrected MNH. FKM's higher post-saccadic velocity floor traded recall for precision. The remaining changes had no measurable effect on these recordings. FKM's improvement over MNH is therefore a filter-alignment correction that MNH users can apply directly, plus a segmentation rule for multi-peak saccades.
Read the full abstract

Materials Science

Giant gas permeance through a nanoporous crystalline 2D polymer

Chemical separation processes represent one of the largest contributors to industrial energy consumption, underscoring the need for the most efficient separation strategies. Many separation processes rely on selective transport through porous media. In this context, two-dimensional (2D) materials offer a promising route to atomically precise porosity beyond the reach of traditional top-down fabrication, with the potential to enhance separation performance by increasing flow. Here we report measurements of gas transport through a one-molecule-thick layer of suspended, nanoporous poly(carboxyfantrip), a 2D polymer that has identical, nanometer-sized pores that cover nearly a quarter of the surface. This combination of ultimate thinness with high areal pore density gives normalized gas flows (permeances) that approach the effusion limit, three to four orders of magnitude above gas-sieving graphene membranes. We find that gases spanning nearly two orders of magnitude in molecular weight (H$_2$, He, N$_2$, CO$_2$, and SF$_6$) obey Graham's law and approach the ideal limit of non-interacting ballistic transport, despite strong, species-dependent gas-pore interactions. Simulations reproduce the measured permeances within experimental uncertainty and provide an atomistic picture of the transport process. This work demonstrates the potential of synthetic crystalline 2D polymers for chemical separation, water filtration, and quantum sensing.
Read the full abstract

All authors Jonathan D'Emidio, Daniel J. Murray, William B. Thompson, Jarvis C. Humlick, K. C. Schwab, Adrian Del Maestro, Benjamin T. King

Computer Vision

InfCoiL: Coordinated Planner-Controller Learning for Closed-Loop Physics-Based Human-Object Interaction

Text- and goal-conditioned human-object interaction (HOI) provides an intuitive interface for specifying high-level intents in physics-based character animation. Mapping such intents to physically plausible, contact-rich behaviors naturally calls for a generative planner and a physics-based controller, yet existing planner-controller systems either execute plans open-loop or close the loop while keeping the planner fixed, so execution feedback never reshapes what is planned. We present InfCoiL, a closed-loop framework for physics-based HOI from coordinated planner-controller learning. InfCoiL couples a morphology-aware flow-matching planner with a unified multi-morphology controller through a shared morphology interface in an autoregressive plan-and-execute loop, and alternately fine-tunes both components with reinforcement learning from the same physical rollouts. To optimize the flow-matching planner online, we introduce Semantic Flow Policy Optimization (SFPO), which stochasticizes selected denoising transitions with channel-wise semantic covariance over heterogeneous HOI features and assigns group-relative, sequence-level rewards from complete closed-loop executions. On held-out OMOMO sequences, InfCoiL outperforms open-loop hierarchical baselines in goal-reaching task success and physical tracking accuracy. Project page: https://infcoil.github.io/
Read the full abstract

All authors Yude Zou, Yifei Yao, Yiwei Hao, Hanqing Wang, Xing Gao

Robotics

Breaking survivorship bias in driving world models

World models allow autonomous vehicles to imagine the consequences of candidate actions before execution, but their reliability depends on the futures represented in their training data. We identify a survivorship bias in driving world models: because the futures recorded in prevailing driving datasets almost always end safely, existing models often imagine benign outcomes even when a collision is imminent. We introduce a neutral world model (NWM) that learns both routine and safety-critical futures from a mixture of routine driving and real accident videos with a bounding-box-guided objective. The NWM more faithfully anticipates imminent collisions, while maintaining a low false collision rate and competitive generation quality in routine driving. We further use the model as an imagination engine for decision-making. Given multiple ego trajectories, the NWM predicts the future associated with each candidate, and a reward function evaluates these imagined outcomes before selecting an action. On challenging NAVSIM scenarios, this procedure selects the better trajectory more reliably than the baseline world model. Zero-shot experiments with real-world pedestrian dummy crossing data collected using an autonomous research vehicle further demonstrate transfer to unseen sensor observations. These results show that correcting survivorship bias enables world models to more faithfully reason about futures and support safety-critical decision-making in autonomous driving.
Read the full abstract

All authors Zihao Sheng, Zilin Huang, Yuhao Luo, Zhengyang Wan, Boyue Wang, Soyoung Ahn, David A. Noyce, Sikai Chen

Machine Learning

Detecting Common Golfing Faults Using AI and Video-Derived Biomechanical Features

Understanding how an athlete moves can improve both technique and safety, yet outside the laboratory technique is still assessed by eye. We present a pipeline that detects named golf swing faults from ordinary video, without markers or calibration. Instead of learning end to end, it tracks body pose and the clubhead in a body-centered frame, and classifies the biomechanical quantities a coach would measure, such as the height of the clubhead above the swing plane, with shallow boosted trees. We demonstrate the approach on two common faults, coming over-the-top (OTT) and early-extension (EE). On 160 swings and 143 swings from public sources it detects OTT and EE with a cross-validated ROC-AUC of 0.975 and 0.911 respectively. Because every feature is a biomechanical quantity a coach already uses, measured at a known phase of the downswing, each decision can be traced to the signals and the moments of the swing that drove it. The signals the model relies on most are the same ones a coach checks when diagnosing these faults. The pipeline processes phone video on a consumer laptop in seconds, so it could feasibly be developed into an app for golfers of all abilities and extended to other sports in which technique is judged against a biomechanical reference.
Read the full abstract

Computers and Society

Rules, Choice, and Chance: A Welfare Evaluation Framework for Student-School Matching

Centralized student-school matching systems, adopted in the United States, Chile, the Netherlands, and the United Kingdom, assign students to schools based on their submitted preference rankings and schools' priority policies. Yet welfare in these systems is typically measured through a narrow lens: a few aggregate statistics published by education authorities, such as the share of students unmatched or assigned to a top-ranked choice. A single aggregate cannot reveal how outcomes vary across students, or why. Welfare depends on three interacting forces: the rules policymakers set (algorithm, tie-breaking, list length limits); the choices students make (which and how many schools to rank); and chance (lottery draws and students' residential location). A rule that reduces unmatched rates may simultaneously lower top-ranked assignment rates for specific groups, and evaluating such trade-offs in advance is challenging because individual-level preference data are rarely available. We present Trace (Trade-offs in Rules, Agency, Chance, and Equity), an open-source framework that synthesizes individual-level preference profiles from aggregate data using mixtures of Mallows models and evaluates alternative rules across multiple welfare metrics, disaggregating outcomes by choice and chance. We apply Trace to two real-world systems: NYC high school matching, where we study minimum list length requirements, and Chilean school assignment, where we compare the current multiple tie-breaking rule with a counterfactual single tie-breaking rule. Both reveal substantial, heterogeneous welfare trade-offs, underscoring the value of simulation-based evaluation before implementing school choice reforms.
Read the full abstract

All authors Reetahan Mukhopadhyay, Clara Strasser Ceballos, Augustin Plantureux, Christoph Kern, Julia Stoyanovich

Data Structures and Algorithms

Induced-Minor-Free Graphs Admit a Star-Based Separator of Size $\tilde{O}(\sqrt{n})$

Given an $H$-induced-minor-free graph $G$ with $n$ vertices, we construct a balanced separator of $G$ consisting of $\tilde{O}(|V(H)|\sqrt{n})$ star subgraphs. Our result is an induced-minor-free analog of the famous separator theorem for $H$-minor-free graphs by Alon, Seymour, and Thomas [J. Amer. Math. Soc., 1990]. A key idea in our construction is a new decomposition of $H$-induced-minor-free graphs, analogous to the cop decomposition for $H$-minor-free graphs by Andreae [J. Combin. Theory Ser. B, 1986]. We also discuss several algorithmic applications of our result.
Read the full abstract

All authors Arnold Filtser, Nikolai Karol, Hung Le, Florian Lehner

Machine Learning (Statistics)

Amortized Bayesian Experimental Design with In-Context Knowledge Conditioning

Amortized Bayesian experimental design (BED) enables real-time design strategies by shifting acquisition cost offline, but existing methods remain tied to the prior and task distribution they were trained on and cannot exploit additional information available at deployment. We introduce in-context amortized BED with unified knowledge encoding (IMBUE), a BED framework that admits external knowledge at deployment through a unified in-context interface. IMBUE accepts two forms of such knowledge without retraining, namely user-specified prior knowledge and observations from previous instances of the experiment. Both are encoded as additional input tokens and processed in-context with the accumulated observations by a shared Transformer, and a learned reliability filter scores each external token against the observations and excludes inconsistent ones. Across four benchmarks, IMBUE accelerates early-stage information acquisition when the external knowledge is reliable, and remains close to the no-external baseline when it is not.
Read the full abstract

All authors Zhanghu Zhao, Daolang Huang, Xinyi Wen, Samuel Kaski

High Energy Physics (Phenomenology)

Next-to-leading-order QCD corrections to radiative decays of fully heavy tetraquarks into quarkonia

Within nonrelativistic QCD factorization, we calculate the next-to-leading-order (NLO) QCD corrections to the decays of compact $T_{4Q}=QQ\bar Q\bar Q$ states ($Q=c,b$) into an $S$-wave quarkonium and a photon. Conservation of charge conjugation allows $T_{4Q}^{0^{++}}$ and $T_{4Q}^{2^{++}}$ to decay into $V_Q+\gamma$, with $V_c=J/\psi$ and $V_b=\Upsilon$, while $T_{4Q}^{1^{+-}}$ can decay into $\eta_Q+\gamma$. At leading order in the heavy quark velocity expansion, we derive the short-distance coefficients for these decay widths in analytic form through $\mathcal{O}(\alpha\alpha_s^5)$, including the interference contribution between the two color configurations of $T_{4Q}^{0^{++}}$. All ultraviolet divergences at NLO are exactly canceled by renormalization. The NLO QCD corrections enhance the decay widths for $T_{4Q}^{2^{++}}\to V_Q+\gamma$ and $T_{4Q}^{1^{+-}}\to\eta_Q+\gamma$ by factors of approximately $2.94$ ($2.33$) and $2.22$ ($1.83$), respectively. The dependence on the renormalization scale $\mu_R$ is reduced at NLO in most channels, although it remains sizable. Based on the NLO branching fractions, we estimate yields of order $10^6$–$10^7$ for $T_{4c}\to J/\psi+\gamma$ and $10^4$ for $T_{4c}\to\eta_c+\gamma$ in $pp$ collisions at $\sqrt{s}=13~\mathrm{TeV}$ with an integrated luminosity of $100~\mathrm{fb}^{-1}$. The $J/\psi+\gamma$ channels can be investigated through $J/\psi\to\ell^+\ell^-$ in the $\ell^+\ell^-\gamma$ invariant-mass spectrum, complementing di-charmonium searches.
Read the full abstract

Computer Vision

Reaching Is Not Finding: Map-Anchored Visual Prompting for the Last Mile of Aerial Vision-Language Navigation

Aerial vision-language navigation (VLN) asks a drone to find an object described in natural language. The description usually locates the object relative to named landmarks, such as streets and buildings. We observe that current navigators often bring the drone close to the target and then stop at the wrong place: they reach the target without finding it. This gap is shared by navigators of very different designs, from imitation learners to a fine-tuned 7B vision-language model. A simple test reveals the cause. When we replace the target object in the instruction with another one next to the same landmarks, no navigator goes to it, and most stop where they did before. Navigators learn to reach the landmarks, not to identify the described object. We therefore propose MAP-VP, a plug-and-play module for the last step. After any navigator stops, MAP-VP draws the landmarks named in the instruction onto the drone's current view, using the landmark outlines that the benchmark provides on a map. A fine-tuned 4B vision-language model then locates the target relative to these landmarks. The training data are generated automatically from the benchmark annotations. Neither drawing the landmarks nor fine-tuning helps much alone, but together they are highly effective. Without retraining any navigator, MAP-VP significantly improves all six navigators we test on CityNav. It roughly doubles the success rate of the two strongest trained navigators, raising HETT from 25.9% to 54.6% and FlightGPT from 24.2% to 48.9%, whereas CityNav's official detector-based last step adds less than two points to HETT.
Read the full abstract

All authors Ningbo Li, Kebin Liu, Jingcan Chen, Weichen Zhang, Rong Ding, Yunhao Liu, Mo Li

Robotics

Review the Episodes, Refine the Lessons: Simple Memory Evolution for Embodied Agents

Recent studies of embodied agents based on frontier multimodal large language models (MLLMs) develop harness engineering around the MLLM, such as harnesses with post-trained policies and recursive self-improvement loops. Yet whether these additions exceed the native capability of the MLLM remains unexplored. We introduce SimplEvo to explore how far an MLLM can improve under direct control when it retains its execution experience in self-written text memories. It requires no weight updates, post-trained policies, or privileged information. The MLLM writes a coordinate-free task memory once from a single demonstration. Under a strict train-test protocol, it then creates a bounded working memory on training episodes only and refines it after each episode with lessons on procedure, completion, and recovery. Both memories stay fixed at test time. On EBench, SimplEvo lifts the success rate (SR) of GPT-6-Astra under direct control from 46.7% to 81.7%, and tops the RoboDojo leaderboard with 69.63% SR and a score of 75.32. Compared with memories that also store past visual frames, SimplEvo reaches a similar SR with text memory alone while using fewer tokens. With its memories fixed, SimplEvo also generalizes to unseen layouts and randomized scene configurations. Moreover, the working memory transfers across models: written by GPT-6-Astra, it can be reused as is by weaker MLLMs and improve their performance without any adaptation of their own.
Read the full abstract

All authors Yifei Yao, Yude Zou, Xing Gao, Guanbo Jiao, Yukai Wang, Jianming Ma, Weinan Zhang, Chunhua Shen

Data Structures and Algorithms

Amortized Cell-Probe Lower Bounds for Dynamic Boolean Problems

We prove the first unconditional amortized $\Omega\big((\log N/\log\log N)^2\big)$ cell-probe lower bound for a Boolean dynamic problem. For insert-only multiphase Disjointness with $N$ query sets over $[N]$ and word size $w=\Theta(\log N)$, a fixed distribution of interleaved insertions and queries forces this expected average cost per operation, excluding preprocessing. The bound allows sufficiently small constant error for each query of every fixed legal sequence and imposes no restriction on individual operation costs. It transfers to incremental directed reachability, incremental subgraph connectivity, and fully dynamic triangle detection. We also prove $t_q=\Omega\big((\log N/\log(t_uw))^2\big)$ for amortized insertion time $1\le t_u=N^{o(1)}$ and expected query time $t_q$, when the number of insertions before the final query is not announced in advance. We reach the quadratic-logarithmic scale with bounded error by combining interleaving with the 2.5-round communication framework [Ko26a, Ko26b] and controlling errors among accepted simulations.
Read the full abstract

Machine Learning

ConvLUT: Rethinking Convolutional Architectures for Ultra-Low-Latency LUT-Based Inference

Lookup-table (LUT)-based neural networks (NNs) offer a compelling path to ultra-low-latency inference on FPGAs. However, existing LUT-based neural networks have largely been restricted to MLP-style topologies, leaving structured vision and audio tasks underexplored. We present ConvLUT, a family of fully unrolled LUT-mappable convolutional architectures targeting the MLPerf Tiny benchmarks. It is comprised of a standard convolutional model built around a hierarchical tree of LUTs and a depthwise separable model following the MobileNet design philosophy. These architectures are enabled by four key contributions: a unified LUT-based convolutional framework supporting standard, depthwise, and pointwise convolutions under explicit fan-in constraints; a LUT-tree that addresses the $k^2c$ receptive field dimensionality bottleneck; a channel-partitioned LUT pooling layer that replaces conventional spatial pooling; and a hardware-aware iterative pruning strategy for stable convergence at scale. Evaluated on CIFAR-10, Visual Wake Words, and Keyword Spotting, ConvLUT achieves competitive accuracy with fully unrolled inference latencies below $\sim 50$ ns. On CIFAR-10 it reaches at least a $8844\times$ latency reduction over non-logic-based prior implementations, and outperforms existing MLP-style LUT-based models on the same task by at least $20$ pp in test accuracy.
Read the full abstract

All authors Marta Andronic, Olivia Weng, Danial Zuberi, Ryan Kastner, George A. Constantinides

Cryptography and Security

On the Relative Security of Prompt Injection: Mechanised Definitions, Defences and Verdicts for LLM Agents

An LLM agent that reads untrusted content, such as an email or a web page, may follow instructions hidden in it. Noninterference is too strong for such agents, since a useful agent lets untrusted content influence what it does and may disclose the user's data by design, and declassification needs a release policy written for each task. We instead state security relative to a reference agent, which runs the same plan under the same runtime monitor with a model that never obeys untrusted content. An agent is relatively secure if every leak it has is a leak the reference already has. Our framework, PlanTERS, instantiates relative security, a notion from speculative execution, with a small agent language in which the attacker chooses how the model follows instructions, a composition with runtime monitors, and attacker models that differ in what they observe. A fence theorem shows that revoking injectable tools at the first untrusted read, and never granting them back, makes every plan relatively secure at the event level on finite runs, for every model behaviour in which each injection is a single call and the model obeys on a schedule fixed in advance; with a fairness condition, this extends to possibly infinite runs. We prove or refute relative security for agents with fences, budgets, approvals, loops and dual LLMs, in Isabelle/HOL.
Read the full abstract

Instrumentation and Detectors

A 40 Gbit/s FPGA-based RoCEv2 transport for detector readout: development and performance

Recording the complete input of a detector trigger allows trigger algorithms to be replayed and studied offline. For the BESIII main drift chamber (MDC), this requires moving tens of Gbit/s from trigger FPGAs into server memory without loss. Remote direct memory access over Converged Ethernet (RoCEv2) lets commercial network adapters write such data into host memory. Detector readout also needs record boundaries that the receiver can check, replay from a verified copy, and host buffers that are reused only after consumption. We present a 40 Gbit/s FPGA RoCEv2 transport that provides these functions. Records are aggregated into CRC32C-checked blocks, and unacknowledged data are kept in DDR memory and verified before replay. Acknowledgment-aware retention writes to DDR only the data that are not acknowledged in time. The transport runs on a ZCU102 board and on the 9U VME Trigger Core Board (TCB). With records of 128 B to 4 KiB it comes within 0.4% of the packetization limit of the link, delivering 36.9 Gbit/s of payload at a path MTU of 1024 B, and its block latency is set by the time needed to fill a block. On a direct link, acknowledgment-aware retention cuts DDR writes for large blocks from 106% to 0.03% of the payload, and injected DDR access errors stay within their queue pair. On the TCB, three 16.4 TB transfers through two switches lost at most five packets each, all recovered by replay. An installation at the MDC is under test.
Read the full abstract

All authors Sheng Dong, Jingzhou Zhao, Haoxin Wang, Tingxuan Zeng, Fei Li

Machine Learning

Data-informed Adaptation of Foundation Models for Causal Effect Estimation

Due to the fundamental problem of causal inference – the impossibility of observing counterfactual outcomes – there is always a gap between the training goal (outcome, propensity) and the inference target (CATE, ATE). Traditional causal effect estimation revolves around setting assumptions, selecting identification strategies, and choosing causal estimators, while more recently, Causal Foundation Models (CFMs), built on Prior-Data-Fitted Networks (PFNs), ease this process by pretraining on massive synthetic prior datasets. Despite their success, such amortization makes causal inference less transparent, obscuring which assumptions the model implicitly relies on for a given dataset. The high diversity of real-world data-generating processes (DGPs) and Structural Causal Models (SCMs) makes it hard to know, a priori, which causal inference strategy is compatible with a given dataset. In this work, we focus on this compatibility between an observational dataset and method-specific assumptions, and find that geometric characteristics of data embeddings – specifically, the variogram – provide a useful signal of whether a given causal estimation strategy is appropriate. We develop MoE-Learner, a mixture-of-causal-experts framework that embeds a dataset under multiple assumption-specific views and uses the variogram as a counterfactual-free router to assess each expert's validity. Rather than eliminating causal assumptions, this shifts their validation from a fixed, a priori choice to a criterion that can be checked directly from data geometry. MoE-Learner achieves the best average rank on benchmark datasets (IHDP, ACIC, Lalonde). Being orthogonal to (causal) foundation models, our method builds a bridge from data characteristics to causal estimation strategy rather than proposing another CFM.
Read the full abstract

Computer Vision

AgentCT: A Vision-Language Agent for Universal CT Imaging

Computed tomography (CT) restoration methods often use a single network for low-dose CT (LDCT), sparse-view CT (SVCT), and metal artifact reduction (MAR), despite the distinct physics of these degradations. We propose AgentCT, a vision-language agent that identifies the degradation and routes images to task-specific restoration models. A LoRA-adapted Qwen3-VL-4B controller predicts degradation type and severity, selects a specialist, and reassesses the restored image. The LDCT specialist uses an image-domain diffusion model, whereas the SVCT and MAR specialists combine projection-domain correction, data consistency, differentiable filtered back-projection, and image refinement. We evaluated the controller and specialists separately on patient-disjoint Mayo 2020 abdominal CT data across three severity levels for each degradation. The controller achieved 95.7% accuracy for degradation type and 93.9% for severity, with a 100% JSON parse rate. Using ground-truth degradation labels for routing, AgentCT achieved an overall PSNR/SSIM of 46.18 dB/0.984, compared with 44.93 dB/0.976 for AMIR and 44.29 dB/0.979 for TAMP, and attained the highest PSNR in all nine settings. For SVCT, its dual-domain specialist achieved 46.04 dB with 4.2 million parameters, compared with 44.43 dB for single-task AMIR and 43.57 dB for all-in-one AMIR with 23.5 million parameters. These results support degradation-specific CT restoration with controller-based model selection; evaluation under automatic routing and on independent clinical data remains necessary. Code is available at https://github.com/hao1635/AgentCT.
Read the full abstract

All authors Zhihao Chen, Qi Gao, Chuang Niu, Qiegen Liu, Yi Zhang, Hongming Shan

Computation and Language

Does Merge Form a Hopf Algebra? Coproducts, Traces, and the Prospects of Algebraic Minimalism

Marcolli, Chomsky, and Berwick (2025; henceforth MCB) formulate the Merge operation of Minimalism on a Hopf algebra of Connes-Kreimer type, and claim that language and physics obey the same algebraic structure. This paper examines the premise of that claim: whether Merge on workspaces can be described as a Hopf algebra. We first survey five coproducts for coassociativity and a connected grading: the Connes-Kreimer coproduct $\Delta_\rho$, MCB's two coproducts $\Delta_d$ and $\Delta_c$, the contraction coproduct of Marcolli and Skigin (2025), and the coproduct of Xiao (2026). $\Delta_d$ is not coassociative. For $\Delta_c$, MCB gives no rule for the labels of trace leaves in trees that already contain trace leaves; under the literal extension it is not coassociative, but once a rule and a grading are supplied it yields a graded connected Hopf algebra. The coproduct of Xiao is coassociative, and so is that of Marcolli and Skigin once its definition is extended to trees with trace leaves. These properties do not tell the coassociative coproducts apart; what distinguishes them is the representation that movement leaves behind. We then review how generative grammar has treated moved elements, and confirm that semantic interpretation requires two things of the representation that results from movement. Retention requires that the structure the lower position had before movement be recoverable; Identifiability requires that, in interpreting the lower position, it be determinable which upper element, if any, corresponds to it. None of the five coproducts meets both conditions: Retention is met only by $\Delta_c$, and there only in the form of a label, and Identifiability by none. In light of the results on coassociativity, we also reassess the Birkhoff factorization that MCB employs at the syntax-semantics interface. Finally, we indicate at least three directions for algebraic Minimalism.
Read the full abstract

2611.11502 · 8 Nov · 26 pages. v2: substantially revised. Corrects v1's claim that the contraction coproduct cannot be coassociative while retaining the labels of traces; adds the coproducts of Marcolli and Skigin (2025) and Xiao (2026); restates the requirement of semantic interpretation as two conditions on the representation that results from movement (Retention and Identifiability), in light of Collins (2026)

Information Theory

Compression of Sources with Memory and Private Random Access

Private local decoding imposes a stringent constraint on compression: the codeword bits used to recover a source symbol must reveal nothing about the remaining symbols beyond what the recovered symbol itself reveals. For memoryless sources, this constraint is known to be compatible with entropy-rate compression. We show that the same holds for any stationary ergodic source over a finite alphabet. Specifically, every rate above the entropy rate is achievable with exact privacy and error-free local decoding on a set of source sequences whose probability tends to one. In contrast, we exhibit non-ergodic sources for which private local decoding incurs a strictly positive rate penalty.
Read the full abstract

Computer Vision

ForeAct: Geometry-Guided Action Learning from World Representations

Robotic manipulation requires representations that capture both scene geometry and its evolution during interaction. Existing approaches have largely treated these two aspects separately. World action models capture such dynamics by coupling action prediction with future video generation, but their reliance on iterative diffusion or rectified-flow inference leads to high computational cost, while their representations lack explicit spatial structure. Explicitly modeling scene geometry typically requires auxiliary modules that increase both parameter count and computational cost. Conversely, geometry-centric policies primarily recover observed spatial structure and often lack explicit supervision for anticipating future geometric changes during interaction. To address these limitations, we introduce ForeAct, which repurposes a pretrained video world model as a single-pass representation backbone for low-latency control. ForeAct directly processes the clean latent of the current observation in one forward pass, eliminating future-frame synthesis and iterative denoising while retaining the backbone's pretrained spatiotemporal priors. Rather than introducing a separate geometry backbone, ForeAct uses learnable tokens supervised by next-frame relative depth to shape the shared video representation through bidirectional self-attention. This compact interface turns the spatial knowledge already encoded in the backbone into a future-aware representation for action generation, with little parameter overhead. The lightweight geometry readout is used only during training and removed at deployment, enabling low-latency, geometry-aware control without depth input or an external geometry backbone. Without additional large-scale pretraining, ForeAct achieves 96.6% average success on LIBERO, 63.5% on SimplerEnv-Bridge, and a 74.4% success rate across 90 real-world rollouts.
Read the full abstract

All authors Wei Lu, Jiaxi Zhang, Yunheng Wang, Taowen Wang, Yuetong Fang, Zhenhui Zhang, Shuning Zhang, Zizhao Yuan, Yixiao Feng, Zecui Zeng, Renjing Xu