04:36 <Ada> arXiv.org | AI-Based On-Board Maritime Object Detection for Earth Observation Payload Data Reduction on Versal Embedded Hardware 4h ago · 3 min read Very-high-resolution Earth-observation satellites acquire more data than they can store and downlink, while in maritime surveillance the vessels cover a tiny fraction of each scene. We study onboard vessel detection as a way to select what is downlinked, which reduces the data according to its content rather than coding every pixel; it is complementary to... Read full story

04:36 <Ada> arXiv.org | Sera: Semantic Representation Aggregation for Reliable and Interpretable Battery Health Forecasting 4h ago · 3 min read Battery state of health (SoH) forecasting is important for battery management, but remains challenging due to nonlinear degradation and heterogeneity across batteries. Existing data-driven approaches primarily use temporal models to learn from numerical battery time series, and higher-level degradation characteristics are often not explicitly represented.... Read full story

04:36 <Ada> arXiv.org | Constitutional Gating and Deterministic Recovery for Multi-Agent LLM Negotiation: Ablations Against a Stateful Adversarial Gatekeeper 4h ago · 4 min read Multi-agent LLM systems negotiating with a stateful counterpart waste model calls in three ways: polite loops that never meet the counterpart's hidden acceptance condition, malformed outputs that trigger retries, and compliance deadlocks in which the counterpart demands something the agent must refuse. We study a three-part control stack - a 5-Pillar runt... Read full story

04:36 <Ada> arXiv.org | Rethinking Contrastive Loss in CLIP Post-training: A Complementary Framework with Frozen Text Encoder 4h ago · 3 min read CLIP serves as a foundational vision-language model and the de facto vision encoder for downstream VLMs such as LLaVA. Post-training offers a lightweight route to refine CLIP, but recent work argues that the standard contrastive loss is unsuitable for post-training due to catastrophic forgetting under small batches, motivating designs that abandon the con... Read full story

04:36 <Ada> arXiv.org | Higher-Order Action Supervision Makes A Strong Policy Class 4h ago · 3 min read Modern data-driven decision-making methods, such as imitation learning (IL) and reinforcement learning (RL), have achieved great success in solving many complex tasks. However, these methods often suffer from serious control instability and robustness issues when applied in real-world applications such as robotics and autonomous driving, posing notable ch... Read full story

04:36 <Ada> arXiv.org | LinSlot: Exploiting Linear Representation hypothesis for unsupervised attribute discovery from slot based object representation 4h ago · 3 min read This paper studies the problem of learning disentangled representations of objects and their attributes from raw, unstructured image data. Slot-based methods have shown considerable success in unsupervised learning of object representations from images. Block-slot attention-based methods extend this framework to attribute representations by assuming a uni... Read full story

04:36 <Ada> arXiv.org | PHRBench: A Behavioral Evaluation of Post-Hallucination Reasoning in LLMs 4h ago · 3 min read Hallucinated information can propagate through multi-stage LLM systems and become part of the context for subsequent reasoning. Existing studies of post-hallucination reasoning (PHR) mainly characterize changes in final outcomes and aggregate reasoning dynamics, leaving how models resolve hallucinated premises at the response level insufficiently understo... Read full story

04:36 <Ada> arXiv.org | Decoupling Logic from Persona: Structural Immunity of Edge LLM Agents to Context Pollution 4h ago · 4 min read Edge language-model agents must hold a persona and reason correctly at once, in one context window that fills with history and persona instructions. We study what happens to the logical part of such an agent when that history is long, misleading and persona-heavy (persona-logic interference), and present a Decoupling Architecture (AO-DA) that separates lo... Read full story

04:36 <Ada> arXiv.org | Work While They Sleep: Exploiting Evaluation Latency for Fully Bayesian Optimization 4h ago · 3 min read Black-box optimization problems are ubiquitous across science and engineering, often dealing with expensive objective functions. This objective latency has two consequences during optimization: (i) the objective evaluation dominates execution time, and (ii) sample-efficient algorithms are crucial to accelerate development and avoid wasting resources. Baye... Read full story

04:36 <Ada> arXiv.org | TRIAGE: Direction-Aware Mismatch Stabilization of Native NVFP4 Reinforcement Learning 4h ago · 3 min read Low-precision execution can substantially accelerate reinforcement learning (RL) for large language models, but discrepancies between learner and sampler execution can destabilize policy optimization. In this paper, we characterize the interaction between mismatch and the policy-gradient direction, distinguishing locally amplifying from contracting update... Read full story

05:01 <Ada> arXiv.org | asdex: Automatic Sparse Differentiation in JAX Adrian Hill, Guillaume Dalle · 5h ago · 3 min read Many tasks in scientific computing and machine learning require the Jacobian or Hessian matrix of a function. Automatic differentiation (AD) computes these derivatives to machine precision, but materializing a dense $m \times n$ Jacobian requires $n$ forward-mode or $m$ reverse-mode AD passes, one per column or row. For a large class of functions, each ou... Read full story

05:01 <Ada> arXiv.org | Rehearse Everything, Remember Nothing: Attic-KV Rehearses What Will Be Read Zhiyun Shi · 5h ago · 3 min read Many key-value (KV) caches are compressed before anyone knows what will be asked of them: a document cached for retrieval, a prompt prefix shared across requests, the memory of a long conversation. The prevailing approach scores KV entries by rehearsal: the model rereads the context and keeps the entries it attends to, assuming that the more completely a ... Read full story

05:01 <Ada> arXiv.org | Efficient quadratic entropy with distance sketches Steve Huntsman · 5h ago · 2 min read We detail scalable methods for approximating the quadratic entropy $p^T d p$ for arbitrary distributions $p$ and common distances $d$ of negative type. We focus on the Euclidean and spherical geodesic cases, which both use random feature embeddings and projections to dramatically improve computational complexity within a simple framework. Amortization of ... Read full story

05:01 <Ada> arXiv.org | Beyond QAOA: A Review of AI and Quantum Computing for Adaptive Combinatorial Optimization Hoong Chuin Lau · 5h ago · 3 min read Near-term quantum approaches to combinatorial optimization are limited by qubit counts, circuit fidelity, sampling cost, and the difficulty of encoding constraints, while machine learning is increasingly used to configure and control quantum optimization workflows. We call such workflows adaptive: decisions conventionally fixed in advance, from formulatio... Read full story

05:01 <Ada> arXiv.org | Smoothing the Top-k Exposure Boundary for Sparse Mixture-of-Experts 5h ago · 3 min read Sparse Mixture-of-Experts models scale parameter capacity efficiently while maintaining a fixed compute budget per token. However, traditional training paradigms enforce a static choice of top-$k$ experts, which converts a continuous routing distribution into a rigid step function. This constraint introduces a brittle boundary where highly competitive exp... Read full story

05:01 <Ada> arXiv.org | Estimating Great Expectations Under Autoregressive Language Models with Potentials Francesco I. Re, Shubhangi Ghosh, Tim Vieira, Ryan Cotterell · 5h ago · 2 min read Many applications of language models hinge not on individual samples but on the expectation of a test functional under the model. Estimating such expectations reliably can be computationally expensive. In this paper, we show how to make estimation more efficient by exploiting the next-token conditional probabilities which are available as a by-product of ... Read full story

05:01 <Ada> arXiv.org | Dissecting Representation Structure in Vision Transformers: A Rigorous Architectural Study Kim-Cuc Nguyen, Ngai-Man Cheung · 5h ago · 3 min read Representation structure is crucial for understanding Vision Transformer (ViT) architectures and their generalization behavior. However, prior studies neither isolate nor analyze module-level features nor investigate how their interactions contribute to performance estimation. In this work, we conduct the first rigorous analysis of feature information acr... Read full story

05:01 <Ada> arXiv.org | Democratizing MoE inference on commodity GPUs with CoMoE 5h ago · 3 min read Deploying Mixture-of-Experts (MoE) models relies heavily on Expert Parallelism, which generates intense inter-GPU communication. Consequently, state-of-the-art inference systems require high-bandwidth, P2P interconnects (e.g., NVLink) in datacenter GPUs to handle massive token routing, making deployment prohibitively expensive. Consumer GPUs offer compara... Read full story

05:01 <Ada> arXiv.org | Workhorse: Learning Robust Whole-Body Humanoid Loco-Manipulation from Human Data 5h ago · 3 min read Humanoid robots still struggle to plan contact-rich whole-body manipulation from egocentric RGB and proprioception. Workhorse learns such manipulation from robot-free human demonstrations. A visual planner predicts five-link targets: the poses of the torso, both wrists, and both feet. A reinforcement-learning whole-body tracker follows them on the robot. ... Read full story

05:01 <Ada> arXiv.org | Explicit Asymptotic Bounds for Sequential Calibration Beyond $T^{2/3}$ Eric Dai, Maxwell Fishelson · 5h ago · 3 min read Probability forecasts are calibrated when predicted probabilities match empirical outcome frequencies: among events assigned a probability $p$, we'd hope that the fraction of positive outcomes is close to $p$. We study the problem of sequential forecasting of binary outcomes. The classical $O(T^{2/3})$ bound on expected cumulative $\ell_1$-calibration err... Read full story

05:19 <Ada> arXiv.org | BiGym 2.0: Benchmarking Learned and Agent-Developed Policies for Humanoid Household Manipulation 5h ago · 3 min read Humanoid household manipulation requires the arms to act while the body balances, steps and changes posture. We present BiGym 2.0, an adaptation of BiGym for the Unitree G1 across 20 household tasks using a unified whole-body controller for demonstration and evaluation. The suite provides 60 native human virtual-reality demonstrations per task with synchr... Read full story

05:19 <Ada> arXiv.org | Evolving Hybrid Quantum-Classical Architectures for Image Classification Devroop Kar, Daniel Krutz, Travis Desell · 5h ago · 3 min read Hybrid quantum classical neural networks integrate parameterized quantum circuits (PQCs) with established deep learning architectures, but their performance depends strongly on the choice of quantum circuit architecture, a choice that remains largely manual. Most existing approaches rely on hand-designed or fixed circuit ansätze, requiring circuit structu... Read full story

05:19 <Ada> arXiv.org | Hamiltonian locality testing and certification do not achieve the Heisenberg limit Francisco Escudero Guti\'errez, Junseo Lee, Sebastian Zur · 5h ago · 3 min read We establish lower bounds for Hamiltonian property testing with access to the time-evolution operator but not its inverse. Each experiment may query the time-evolution operator multiple times, and distances between Hamiltonians are measured in the normalized Frobenius norm. In this model, we show that testing whether a Hamiltonian is $k$-local or $\vareps... Read full story

05:19 <Ada> arXiv.org | Sharp Statistical Rates for Asynchronous TD Learning with Markovian Data Yang Peng · 11d ago · 3 min read We study the last iterate of standard tabular temporal-difference (TD) learning from a single trajectory of a finite Markov reward process. For discount factor $γ$, write $H=(1-γ)^{-1}$, and let $μ_{\min}$ and $t_{\operatorname{mix}}$ denote the minimum stationary probability and total-variation mixing time. We prove that last-iterate TD achieves sup-norm... Read full story

05:19 <Ada> arXiv.org | OpenCollab: A Multi-Agent Coding Framework with Programmable Collaboration and Controllable Runtime 5h ago · 3 min read Multi-agent coding systems are designed to tackle complex software engineering tasks through collaboration. However, existing evaluations typically assume configured organizations are followed faithfully, whereas reality differs. This behavioral gap, combined with differences in underlying system components, prevents clear attribution of observed gains. T... Read full story

05:19 <Ada> arXiv.org | Robust LassoNet: Enhancing Feature Selection in Neural Networks via Robust Loss Functions Daniela De Canditiis, Italia De Feis, Paola Stolfi · 11d ago · 3 min read Feature selection in neural networks remains a challenging problem, particularly in the presence of noisy or contaminated data. LassoNet is a recent approach that addresses this issue by combining neural networks with hierarchical sparsity constraints, enabling simultaneous prediction and variable selection. However, its standard formulation relies on the... Read full story

05:19 <Ada> arXiv.org | ThinkNet: Evolutionary Deep Ensembles for Efficient Motor-Imagery EEG Classification Abdul Basit, Saim Rehman, Muhammad Shafique · 5h ago · 3 min read Practical assistive and rehabilitative brain--computer interfaces require subject-independent motor-imagery EEG (MI-EEG) decoders that generalize to new users under limited target-user data and constrained compute. However, held-out-subject performance can be overstated when test-subject information influences preprocessing, model selection, or ensemble s... Read full story

05:19 <Ada> arXiv.org | Trust, but Validate the Instrument: Auditing AI-Generated RTL Verification Plans on Authored Security-Regression Proxies Hang Xiao, Chuhong Xu, Kainan Zhou, Gangzhen Qian, Lu Yi · 5h ago · 3 min read AI-generated RTL verification plans can satisfy a provider schema yet fail at the boundary to trusted execution. We present SecTB-RTL, an auditable framework covering 31 tasks and 124 authored hardware-security regressions. A deterministic non-AI baseline killed 36, 75, and 78 mutants at increasing resource limits. The first confirmatory run (C1-R2) faile... Read full story

05:19 <Ada> arXiv.org | A multimodal large language model for evidence-based autism spectrum disorder screening 5h ago · 3 min read The clinical management of autism spectrum disorder (ASD) faces a bottleneck in early screening, mainly because trained specialists are scarce and conventional assessment tools are subjective. Here, we introduce ASDchat, a multimodal large language model designed for evidence-based ASD screening, which takes video, audio, and dialogue as input. ASDchat ad... Read full story

05:19 <Ada> arXiv.org | Predicting Social Media Engagement using Machine Learning Ritwik Singh, Mayukh Majumdar, Subodha Kumar · 5h ago · 3 min read Social media platforms are popular channels for disseminating information, owing to their large user bases and ease of access. Companies also use social media as an important aspect of the advertising process. By creating high-quality posts, companies can strengthen their engagement metrics and increase their follower count. While a growing body of resear... Read full story

05:20 <Ada> arXiv.org | Nonsmooth Optimization via Orthogonalized Momentum Lexiao Lai, Tianyi Lin, Jiayu Zhang · 5h ago · 3 min read Modern real application problems involve matrix-valued parameters, yet conventional optimizers treat them as vectors, thereby motivating matrix-aware methods that exploit input-output geometry, such as Muon which orthogonalizes the momentum matrices before parameter updates. Its empirical success raises a conceptual question: can orthogonalized momentum r... Read full story

05:20 <Ada> arXiv.org | What Makes Adversarial Examples Transfer Across Deepfake Detectors? Rafael M. Mamede, Pedro C. Neto, Ana F. Sequeira · 5h ago · 3 min read Deepfake detectors remain vulnerable to transfer-based black-box attacks, in which adversarial examples are generated on a source surrogate model and transferred to a target model, unknown to the attacker. Yet how source--target compatibility shapes attack success remains poorly understood. Prior studies evaluate limited detector pools and rarely disentan... Read full story

05:20 <Ada> arXiv.org | Differentiable Hybrid Modelling for Learning and Optimising Chemical Transport Processes from Experimental Data 5h ago · 3 min read Transport models underpin the design and optimisation of chemical processes, yet their constitutive laws are typically hand-picked, their experimental initial conditions are assumed exact, and their computational cost restricts optimisation to a handful of process variables. Here we introduce a differentiable hybrid modelling framework that embeds neural ... Read full story

05:20 <Ada> arXiv.org | Preference Elicitation for Policy Optimization and Application to Aligning Heart Transplantation with Human Values 5h ago · 3 min read Preference elicitation is essential for aligning AI systems with human values. Prior approaches (e.g., for organ allocation) often ask stakeholders to compare the decisions of an algorithm (e.g., patient A vs. patient B). Such a decision-level approach conflates the means with the ends. Instead, we elicit preferences directly over allocation outcomes to l... Read full story

05:20 <Ada> arXiv.org | Enhancing Tabular Learners with Context-Aware Semantic Embeddings G\"unther Schindler, Maximilian Schambach, Johannes H\"ohne · 5h ago · 3 min read While modern tabular learners excel at capturing statistical patterns, they frequently operate in a semantic vacuum, treating textual features as discrete symbols, ignoring the rich semantics inherent in feature names or cell entries. We propose CASE (Context-Aware Semantic Embeddings), a novel framework that bridges the gap between the semantic understan... Read full story

05:20 <Ada> arXiv.org | DRIFT: Derailing Denoising Trajectories of Flow-Matching VLAs with Adversarial Patch Attack Hoseong Tae, Jong-Seok Lee · 5h ago · 3 min read Flow-matching vision-language-action (VLA) models such as pi0 generate robot actions by integrating a learned denoising velocity field, and have been reported to resist adversarial perturbations that readily fool autoregressive VLAs. We show that this robustness is largely illusory: it stems from prior attacks ignoring the multi-step denoising ODE. We int... Read full story

05:20 <Ada> arXiv.org | Particle-based Generalised Stochastic Optimisation Jiechen Jackie Zhang, O. Deniz Akyildiz · 5h ago · 2 min read We develop a class of general diffusion-based stochastic particle optimisation methods for loss functions with intractable gradients. Specifically, we consider problems in which the loss gradient is an integral with respect to a parameter-dependent distribution, a structure that includes training generative models, fine-tuning, and learning latent-variabl... Read full story

05:20 <Ada> arXiv.org | MIRROR: Learning from the Other View for Multi-Modal Reasoning Wen Ye, Yuxiao Qu, Aviral Kumar, Xuezhe Ma · 5h ago · 3 min read Vision-language models (VLMs) can behave very differently depending on whether the same underlying problem is presented textually or visually. This inconsistency suggests that different views expose different reasoning paths and failure modes that standard multimodal post-training does not fully exploit. To study and exploit this phenomenon, we construct ... Read full story

05:20 <Ada> arXiv.org | Medical foundation models converge less under label supervision 5h ago · 3 min read Diagnostic classifiers and imaging biomarkers are fitted on the embeddings of medical foundation models. These models are replaced as new versions appear. This practice assumes that different models represent images alike. We tested this assumption with more than 750,000 images from 14 datasets in 5 imaging modalities, 18 public models, and 101 models tra... Read full story

05:20 <Ada> arXiv.org | Equality, Equity, and Causality in Fairness Research: A Commentary on Cheng (2026) Youmi Suk · 5h ago · 2 min read This is an invited commentary on the Psychometrika focus article "Fairness Issues and Evaluation in Psychometrics and AI/ML: What Can We Learn from Each Field?" by Ying Cheng (2026, doi:10.1017/psy.2026.10110). Cheng offers a systematic comparison between long-standing test fairness and modern algorithmic fairness. Her mapping of the entire testing workfl... Read full story

08:47 <Ada> arXiv.org | H-OPD: Confidence Aware Heterogeneous Multi-Teacher Multimodal On-policy Distillation 8h ago · 3 min read On-policy distillation (OPD) has recently emerged as an effective post-training paradigm by providing supervision on student-generated trajectories. However, existing OPD methods for multimodal reasoning usually rely on a static teacher routing, assigning each sample to a single teacher based on modality or task type. This ignores that visual grounding an... Read full story

08:47 <Ada> arXiv.org | Hierarchical Experimentalist Agents 8h ago · 3 min read Large language models (LLMs) are increasingly used to act in the real world and augment human decision-making, yet most systems rely on parametric knowledge acquired through imitation, optionally improved by post-training, retrieval, or search over fixed data. This paradigm struggles in novel domains or on complex queries that require information unavaila... Read full story

08:47 <Ada> arXiv.org | RoboMME-Interference: Benchmarking Robot Memory Under Interference Soumil Rathi · 8h ago · 3 min read Robots deployed in realistic settings will accumulate experience across many sessions and tasks over their deployment. The robot's tasks may often require it to remember information from multiple sessions ago, making long-context robot memory important for real-world deployments. However, most robot-memory benchmarks today are based on single episodes or ... Read full story

08:47 <Ada> arXiv.org | Decoupling the Declarative from the Procedural in Vision-Language-Action Models 8h ago · 3 min read Deploying generalist robotic agents in the real world requires transferable skills. Specifically, a policy trained to clone a behavior from object-specific demonstrations must generalize beyond that object, otherwise data collection requirements become intractable. Recently, fine-tuning of pre-trained billion-parameter Vision-Language Models (VLMs), initi... Read full story

08:47 <Ada> arXiv.org | TelcoAgent: A Scalable 5G Multi-KPM Forecasting With 3GPP-Grounded Explainability 8h ago · 3 min read Key Performance Measurement (KPM) forecasting is essential for proactive network management of 5G and next-generation telecom networks. However, existing machine learning (ML) approaches face significant limitations in scalability and explainability, restricting their effectiveness in real-world deployments. We propose TelcoAgent, a foundation model-based... Read full story

08:47 <Ada> arXiv.org | Bistable by Construction: Wall-Clock-Calibrated State Monitors Have No Moment-Detection Regime at Agent Cadence Manvendra Modgil · 8h ago · 4 min read Runtime monitors for autonomous agents commonly threshold an accumulated internal state - a behavioural baseline, a drift statistic, or, in our prior work, a modelled affective state. We previously reported a State Saturation Trap: threshold-on-state triggers over a continuous affect engine become near-constant alarms on SWE-bench debugging agents (Modgil... Read full story

08:47 <Ada> arXiv.org | PACT: Privileged Trace Co-Training for Multi-Turn Tool-Use Agents 8h ago · 3 min read Multi-turn tool-use agents must reason, call tools, and adapt to observations across several interaction turns. Post-training such agents is challenging, as reinforcement learning often suffers from sparse rewards and weak credit assignment despite matching the prompt-only inference setting, while supervised fine-tuning on expert traces provides dense pro... Read full story

08:48 <Ada> arXiv.org | Lesion-DDPM: Lesion-Enhanced 3D Diffusion for MS MRI Synthesis 8h ago · 3 min read 3D FLAIR MRI is widely recommended as one of the standard MRI sequences for brain imaging in multiple sclerosis (MS), but publicly available MS datasets remain relatively small and vary across scanners, acquisition protocols, and lesion patterns. This scarcity and variability hinder the development of robust neuroimaging machine learning models and are pa... Read full story

08:48 <Ada> arXiv.org | Agents' Last Exam 8h ago · 8 min read Recent AI systems have achieved strong results on a wide range of benchmarks, yet these gains have not translated into economically meaningful deployment across many professional domains. We argue that this gap is largely an evaluation problem: widely used benchmarks lack sustained performance measurement on real and economically valuable workflows. This ... Read full story

08:48 <Ada> arXiv.org | The Invisible Hand of Physics: When Video Diffusion Models Know More Than They Show 8h ago · 3 min read Modern video diffusion models generate increasingly realistic and temporally coherent videos, motivating their use as candidate world simulators. Yet it remains unclear whether these models internally encode physical structure, or merely reproduce motion patterns seen during training. We study this question by probing video diffusion models along latent t... Read full story

08:48 <Ada> arXiv.org | Unsupervised Identification and Removal of Spurious Correlations During Fine-Tuning 8h ago · 3 min read Fine-tuning a pretrained language model on a curated dataset can produce spurious correlations between the fine-tuning task and unintended latent factors---such as misaligned personas or political slant---that the curation procedure has entangled with the task. The model can latch onto these spurious correlations, leading to bias and reduced out-of-distri... Read full story

08:48 <Ada> arXiv.org | Is TabPFN the Silver Bullet for Insurance Pricing? Bruno Deprez, Wouter Verbeke, Tim Verdonck · 8h ago · 3 min read Modelling claim frequency and severity for non-life insurance pricing predominantly relies on generalised linear models (GLMs), with gradient-boosted machines as the leading machine learning alternative. Tabular foundation models (TFMs) present a fundamentally different inference paradigm. By pre-training on large collections of synthetic datasets, TFMs e... Read full story

08:49 <Ada> arXiv.org | Cross-View Attention Fusion Net: A Prior-Guided Dual-View Representation Learning for Cardiac Output Estimation from Short-Term PPG Signals 8h ago · 3 min read Accurate cardiac output (CO) estimation from photoplethysmography (PPG) is promising for unobtrusive hemodynamic monitoring, but remains difficult since CO is jointly determined by cardiac function and vascular tone. Conventional feature-based models use physiologically meaningful PPG descriptors, yet depend on accurate pulse detection and may miss latent... Read full story

08:49 <Ada> arXiv.org | Stochastic Compositional Optimization via Hybrid Momentum Frank--Wolfe El Mahdi Chayti, Mohamad Fakhouri · 8h ago · 3 min read Stochastic compositional optimization minimizes $F(f(x), x)$ over a compact convex set when the inner map $f$ is available only through noisy samples. We propose Hybrid Momentum Stochastic Frank--Wolfe, a projection-free method for outer functions $F$ that are jointly convex and Lipschitz but not necessarily differentiable, such as maxima, CVaR, and norms... Read full story

08:49 <Ada> arXiv.org | Vision2Code: A Multi-Domain Benchmark for Evaluating Image-to-Code Generation Ajay Vikram Periasami, Junlin Wang, Bhuwan Dhingra · 8h ago · 3 min read We study whether vision-language models (VLMs) can reconstruct structured, graphics-based images by writing executable code to generate them. Existing benchmarks on image-to-code generation are generally limited in the domains they cover, rely on reference source code for evaluation, and do not provide metrics to quantify domain-specific issues. To addres... Read full story

08:49 <Ada> arXiv.org | UxSID: Semantic-Aware User Interests Modeling for Ultra-Long Sequence 8h ago · 3 min read Modeling ultra-long user sequences involves a difficult trade-off between efficiency and effectiveness. While current paradigms rely on either item-specific search or item-agnostic compression, we propose UxSID, a framework exploring a third path: semantic-group shared interest memory. By utilizing Semantic IDs (SIDs) and a dual-level attention strategy, ... Read full story

08:49 <Ada> arXiv.org | Convex Optimization of Spiking Neural Networks with Global and Approximation Guarantees 8h ago · 3 min read Spiking neural networks (SNNs) have been proposed as energy-efficient alternatives to deep neural networks. Their training is non-convex and involves a non-differentiable spike function, so existing methods offer no guarantee of reaching a global optimum. We address this issue by formulating SNN optimization as an equivalent convex optimization problem. S... Read full story

08:49 <Ada> arXiv.org | FactoryBench: Evaluating Industrial Machine Understanding 8h ago · 3 min read We introduce FactoryBench, a benchmark for evaluating time-series models and LLMs on machine understanding over industrial robotic telemetry. Q&A pairs are organized along four causal levels (state, intervention, counterfactual, decision) instantiating Pearl's ladder of causation, and span five answer formats: four structured formats are scored determinis... Read full story

08:49 <Ada> arXiv.org | Fairness-Guided DQN-Based TXOP Control for NR-U/Wi-Fi Coexistence Po-Heng Chou, Yi-Fang Yu, Shou-Yu Chen, Chiapin Wang · 8h ago · 3 min read New Radio operating in unlicensed spectrum (NR-U) and Wi-Fi employ different contention parameters and channel-occupancy durations, which can produce severe coexistence imbalance. This correspondence develops a fairness-guided deep Q-network (DQN) that adjusts the NR-U transmission opportunity (TXOP) using the observed per-node throughput ratio. Two rewar... Read full story

08:49 <Ada> arXiv.org | Occam's Razor is Only as Sharp as Your ELBO Ethan Harvey, Geoff Pleiss, Michael C. Hughes · 8h ago · 3 min read The marginal likelihood, also known as the evidence, is regarded as a mathematical embodiment of Occam's razor, enabling hyperparameter selection that avoids overfitting. The evidence lower bound (ELBO) objective from variational inference has also been used for similar purposes. In this paper, we define a signal-to-noise ratio (SNR) for Bayesian linear r... Read full story

08:50 <Ada> arXiv.org | Component-Adaptive and Lesion-Level Supervision for Improved Small Structure Segmentation in Brain MRI Minh Sao Khue Luu, Evgeniy N. Pavlovskiy, Bair N. Tuchinov · 4d ago · 4 min read Small lesions in brain MRI are hard to segment because they occupy a tiny fraction of the volume and are dominated by background and larger lesions during voxel-wise optimization, so a model can reach a high Dice similarity coefficient (DSC) while missing many of them. We propose CATMIL, a training objective that adds two auxiliary terms to the standard n... Read full story

08:50 <Ada> arXiv.org | MVNN: A Measure-Valued Neural Network for Learning McKean-Vlasov Dynamics from Particle Data Liyao Lyu, Xinyue Yu, Hayden Schaeffer · 8h ago · 3 min read Collective behaviors that emerge from interactions are fundamental to numerous biological systems. To learn such interacting forces from observations, we introduce a measure-valued neural network that infers measure-dependent interaction (drift) terms directly from particle-trajectory observations. The proposed architecture generalizes standard neural net... Read full story

08:50 <Ada> arXiv.org | Locally Private Parametric Methods for Change-Point Detection 8h ago · 3 min read We study parametric change-point detection under local differential privacy, focusing on the offline localization of a single change point in a sequence of independent observations.~In the non-private setting, we derive improved finite-sample accuracy guarantees for a generalized likelihood-ratio estimator using martingale methods.~We establish finite-sam... Read full story

08:50 <Ada> arXiv.org | RanSOM: Second-Order Momentum with Randomized Scaling for Constrained and Unconstrained Optimization El Mahdi Chayti, Mohamad Fakhouri · 8h ago · 3 min read Second-order momentum corrects the lag of momentum behind the current gradient with a Hessian-vector product (HVP), but existing corrections need either a Lipschitz Hessian or extra forward and backward passes at an auxiliary point. We show that randomizing the step length removes this trade-off: if the step is exponentially distributed with mean $η$, a S... Read full story

08:50 <Ada> arXiv.org | Fast Rates for Nonstationary Weighted Risk Minimization Tobias Brock, Thomas Nagler · 8h ago · 3 min read Weighted empirical risk minimization is a common approach to prediction under distribution drift. This article studies its out-of-sample prediction error under nonstationarity. We provide a general decomposition of the excess risk into a learning term and an error term associated with distribution drift, and prove oracle inequalities for the learning erro... Read full story

08:50 <Ada> arXiv.org | HumanX: Toward Agile and Generalizable Humanoid Interaction Skills from Human Videos 8h ago · 3 min read Enabling humanoid robots to perform agile and adaptive interactions has long been a core challenge in robotics. Current approaches are bottlenecked by either the scarcity of realistic interaction data or the need for meticulous, task-specific reward engineering, which limits their scalability. To narrow this gap, we present HumanX, a unified framework tha... Read full story

08:50 <Ada> arXiv.org | A Generalizable Machine Learning Framework for Half-Heusler Thermoelectrics Shoeb Athar, Adrien Mecibah, Philippe Jund · 8h ago · 3 min read Machine learning (ML) has emerged as a powerful approach for accelerating the discovery of thermoelectric (TE) materials, yet many models suffer from limited generalizability. Here we present a robust ML framework for predicting the thermoelectric figure of merit (zT) in half-Heusler (hH) compounds, emphasizing prediction of previously unseen compositions... Read full story

08:50 <Ada> arXiv.org | RPP: A Certified Poisoned-Sample Detection Framework for Backdoor Attacks under Dataset Imbalance 8h ago · 3 min read Deep neural networks are highly susceptible to backdoor attacks, yet most defense methods to date rely on balanced data, overlooking the pervasive class imbalance in real-world scenarios that can amplify backdoor threats. This paper presents the first in-depth investigation of how the dataset imbalance amplifies backdoor vulnerability, showing that (i) th... Read full story

08:50 <Ada> arXiv.org | Learning to accelerate Krasnosel'skii-Mann fixed-point iterations with guarantees Andrea Martin, Giuseppe Belgioioso · 8h ago · 3 min read We introduce a principled learning to optimize (L2O) framework for solving fixed-point problems involving general nonexpansive mappings. Our idea is to deliberately inject summable perturbations into a standard Krasnosel'skii-Mann iteration to improve its average-case performance over a specific distribution of problems while retaining its convergence gua... Read full story

08:50 <Ada> arXiv.org | Probabilistic Multi-Agent Aircraft Landing Time Prediction Kyungmin Kim, Seokbin Yoon, Keumjin Lee · 8h ago · 3 min read Accurate and reliable aircraft landing time prediction is essential for effective resource allocation in air traffic management. However, the inherent uncertainty of aircraft trajectories and traffic flows poses significant challenges to both prediction accuracy and trustworthiness. Therefore, prediction models should not only provide point estimates of a... Read full story

08:51 <Ada> arXiv.org | MetaJoin: Adaptive Join Selections for Heterogeneous Tables Ning Wang, Sainyam Galhotra · 8h ago · 3 min read Joining heterogeneous tables requires different forms of correspondence, including exact agreement, syntactic transformation, semantic similarity, and entity-level matching. The appropriate join method is often unknown in advance, while running every candidate method incurs unnecessary cost. We introduce MetaJoin, a label-free framework for adaptive join ... Read full story

08:51 <Ada> arXiv.org | Infinitesimal Operator World Models: Finite-Sample Guarantees for Offline Reinforcement Learning 8h ago · 3 min read Continuous-time stochastic processes underlie many natural and engineered systems. In healthcare, autonomous driving, and industrial control, direct interaction with the environment is often unsafe or impractical, motivating offline reinforcement learning from historical data. However, there is limited statistical understanding of the approximation errors... Read full story

08:51 <Ada> arXiv.org | Second-Order Optimization with Asynchronous Curvature El Mahdi Chayti, Martin Jaggi · 8h ago · 3 min read Second-order methods make fast progress per iteration, but forming and factorizing the Hessian often dominates their wall-clock time. We study \emph{Split-Client}, a task-parallel design in which a curvature client factorizes Hessians in the background while a gradient client takes cubic-regularized steps with the most recently published matrix, so that c... Read full story

08:51 <Ada> arXiv.org | ABLEIST: Intersectional Disability Bias in LLM-Generated Hiring Scenarios 8h ago · 3 min read Large language models (LLMs) are increasingly under scrutiny for perpetuating identity-based discrimination in high-stakes domains such as hiring, particularly against people with disabilities (PwD). However, existing research remains largely Western-centric, overlooking how intersecting forms of marginalization--such as gender and caste--shape experience... Read full story

08:51 <Ada> arXiv.org | Surrogate Graph Partitioning for Spatial Prediction Yuta Shikuri, Hironori Fujisawa · 8h ago · 3 min read Spatial prediction refers to the estimation of unobserved values from spatially distributed observations. Although recent advances have improved the capacity to model diverse observation types, adoption in practice remains limited in industries that demand interpretability. To mitigate this gap, surrogate models that explain black-box predictors provide a... Read full story

08:51 <Ada> arXiv.org | Approximation of differential entropy in Bayesian optimal experimental design Chuntao Chen, Tapio Helin, Nuutti Hyv\"onen, Yuya Suzuki · 8h ago · 3 min read Bayesian optimal experimental design provides a principled framework for selecting experimental configurations that maximize obtained information. In this work, we consider settings where the likelihood entropy is independent of the design or can be evaluated explicitly, which applies, e.g. to additive noise models with design-independent or explicitly co... Read full story

08:51 <Ada> arXiv.org | UniConFlow: A Unified Constrained Flow-Matching Framework for Certified Motion Planning 8h ago · 3 min read Generative models have become increasingly powerful tools for robot motion generation, enabling flexible and multimodal trajectory generation across various tasks. Yet, most existing approaches remain limited in handling multiple types of constraints, such as collision avoidance, actuation limits, and dynamic consistency, which are typically addressed ind... Read full story

08:51 <Ada> arXiv.org | Physics-inspired Generative AI models via real hardware-based noisy quantum diffusion 8h ago · 3 min read Quantum Diffusion Models (QDMs) are an emerging paradigm in Generative AI that aims to use quantum properties to improve the performances of their classical counterparts. However, existing algorithms are not easily scalable due to the limitations of near-term quantum devices. Following our previous work on QDMs, here we propose and implement two physics-i... Read full story

08:51 <Ada> arXiv.org | Auditing Prompt Injection Defenses Before They Reach Users: What Strong Adversaries Reveal 8h ago · 3 min read Large language models are increasingly deployed in user-facing products that handle sensitive personal data while interacting with untrusted external content. This makes them a prime target for prompt injection attacks, in which an adversary injects malicious instructions in that content to hijack the model. Fine-tuning-based defenses have been proposed t... Read full story

08:51 <Ada> arXiv.org | Fast localized urban microclimate emulator: wind and temperatures in unseen urban morphologies 8h ago · 3 min read Urban microclimate, encompassing wind and temperature fields shaped by building geometry, strongly affects building energy use, pedestrian wind comfort, pollutant dispersion, the urban heat island, and public health. Conventional Computational Fluid Dynamics (CFD) is computationally prohibitive for rapid assessments, while many deep learning approaches re... Read full story

08:57 <Ada> arXiv.org | Understanding Transformer-based Vision Models through Inversion 8h ago · 3 min read Understanding the internal mechanisms of deep neural networks remains a central challenge in machine learning. In computer vision, one promising yet only preliminarily explored approach is feature inversion via inverse networks, which reconstructs images from intermediate representations using trained inverse networks. In this study, we revisit feature in... Read full story

08:57 <Ada> arXiv.org | Data-Driven Stochastic Optimal Control in Reproducing Kernel Hilbert Spaces 8h ago · 3 min read This paper proposes a fully data-driven approach for optimal control of nonlinear control-affine systems represented by a stochastic diffusion. The focus is on the scenario where both the nonlinear dynamics and stage cost functions are unknown, while only a control penalty function and constraints are provided. To this end, we embed state probability dens... Read full story

08:57 <Ada> arXiv.org | Sera: Semantic Representation Aggregation for Reliable and Interpretable Battery Health Forecasting 8h ago · 3 min read Battery state of health (SoH) forecasting is important for battery management, but remains challenging due to nonlinear degradation and heterogeneity across batteries. Existing data-driven approaches primarily use temporal models to learn from numerical battery time series, and higher-level degradation characteristics are often not explicitly represented.... Read full story

08:57 <Ada> arXiv.org | When to Intervene? State-Aware Sparse Manipulation in Federated Reinforcement Learning Shutong Zheng, Sijia Chen · 8h ago · 3 min read Federated reinforcement learning (FRL) enables distributed agents to collaboratively train decision-making policies, but its decentralized training process also exposes global policy learning to Byzantine manipulation. Existing poisoning attacks primarily focus on how to construct malicious updates, while trajectory-level intervention timing remains large... Read full story

08:58 <Ada> arXiv.org | Deflating the Hessian: Rank-4 W4A4 Quantization for Multimodal Diffusion Transformers Shiwen Wang, Pengxiang Zhao, Xiaoming Yuan · 8h ago · 3 min read In diffusion transformers, low-rank branches can mitigate 4-bit weight--activation (W4A4) post-training quantization (PTQ) loss by decomposing each weight into a low-bit residual and a high-precision low-rank component. Existing low-rank PTQ approaches, however, either optimize low-rank compensation and residual quantization separately, often requiring hi... Read full story

08:58 <Ada> arXiv.org | When Do External Priors Help Low-Resource Structured Prediction? A Ranking-Level Alignment Study Qiye Lu, Jiang Ji, Liang Zhang · 8h ago · 3 min read Cold-start credit scoring (scarce labels, immature features, or minimal capacity) motivates injecting external priors (teacher models, LLMs, or domain experts) into training. We study Ranking Prior Alignment (RPA), a model-agnostic objective that aligns a model's ranking distribution to an external prior via a temperature-scaled KL loss with an annealed w... Read full story

08:58 <Ada> arXiv.org | Dynamics as Code: On Model Compression via Dynamic System 8h ago · 3 min read The escalating size of pretrained neural networks has rendered model compression a prerequisite for deployment under stringent memory and compute constraints. With the irrational winding as an example, earlier work introduced a dynamic system (DS) paradigm that reconceptualizes compression as compact weight representation: high-dimensional parameters are ... Read full story

08:58 <Ada> arXiv.org | Measuring and Mitigating Solution Mode Collapse in RLVR 8h ago · 3 min read A language model (LM) can usually answer the same question in more than one way, but reinforcement learning with verifiable rewards (RLVR) is indifferent to which correct answer a model produces. A solution will earn the same reward whether it is the thousandth copy of a familiar answer or one the model has never produced before. Yet, there is potential v... Read full story

08:58 <Ada> arXiv.org | Work While They Sleep: Exploiting Evaluation Latency for Fully Bayesian Optimization 8h ago · 3 min read Black-box optimization problems are ubiquitous across science and engineering, often dealing with expensive objective functions. This objective latency has two consequences during optimization: (i) the objective evaluation dominates execution time, and (ii) sample-efficient algorithms are crucial to accelerate development and avoid wasting resources. Baye... Read full story

08:58 <Ada> arXiv.org | TRACE: Rollout-Guided Quantization-Aware Training for FP4 Reinforcement Learning of MoE Language Models 8h ago · 3 min read Reinforcement learning (RL) for post-training large language models (LLMs) incurs substantial computation and memory overhead during rollout generation, which motivates low-precision rollout for efficient RL training. However, existing FP4 RL methods suffer from a key limitation: they primarily optimize quantization accuracy on the training and rollout pa... Read full story

08:59 <Ada> arXiv.org | TRIAGE: Direction-Aware Mismatch Stabilization of Native NVFP4 Reinforcement Learning 8h ago · 3 min read Low-precision execution can substantially accelerate reinforcement learning (RL) for large language models, but discrepancies between learner and sampler execution can destabilize policy optimization. In this paper, we characterize the interaction between mismatch and the policy-gradient direction, distinguishing locally amplifying from contracting update... Read full story

08:59 <Ada> arXiv.org | Beyond In-Distribution Preservation: Recovering Generalization in Quantized VLAs via Vulnerability-Oriented Tuning 8h ago · 3 min read Post-training quantization has been shown to preserve VLA performance under standard evaluation conditions, but whether it preserves the full-precision model's robustness and generalization remains underexplored. In this study, we systematically study the robustness and generalization of post-quantized VLA policies under environmental disturbances. Empiri... Read full story

08:59 <Ada> arXiv.org | Green-Routed Neural Operators: Physics Determines Where the Network Reads Chenhao Si, Ming Yan · 8h ago · 3 min read We identify a mismatch between the physical role of transport fields in many PDEs and their usual role in neural operators: PDEs use them to select read coordinates, whereas neural operators typically treat them only as input values. We address this mismatch with the Green-Routed Neural Operator (GRNO), which uses the governing equation to determine where... Read full story

08:59 <Ada> arXiv.org | Selection-Based Structured Reasoning: Toward Efficient Multimodal Search Agents 8h ago · 3 min read Multimodal agents commonly generate free-form reasoning before each action. For small models, limited model capacity can result in lengthy reasoning that provides little useful guidance for action generation while incurring substantial inference cost. To address this challenge, we introduce Selection-based Structured Reasoning (SSR), a framework that refo... Read full story

08:59 <Ada> arXiv.org | Exploring More, Reasoning Better: Stepwise Risk-Sensitive GRPO for Diffusion Language Models Yue Yu, Bowen Zuo, David Crandall, Yinglun Zhu, Dongruo Zhou · 8h ago · 3 min read Diffusion large language models (dLLMs) generate text by denoising a sequence or successive blocks, allowing several tokens to be revealed in parallel. Reinforcement learning with verifiable rewards (RLVR) reuses terminal feedback across these decisions, even as their conditioning context changes. We propose stepwise risk-sensitive GRPO (StepRS-GRPO), whi... Read full story

08:59 <Ada> arXiv.org | Useful to Whom? The Easy-versus-Coverage Boundary Moves with the Learner Yangze Liu, Xiao-Long Yin, Zhongyi Han · 8h ago · 3 min read Is the same data equally useful to every learner? The question is whether the value of a training sample belongs to the sample or arises in its interaction with the learner. We test it with a selection experiment. Under a per-class budget, two kinds of samples compete for the slots: easy samples, which a short training run learns first, and coverage sampl... Read full story

08:59 <Ada> arXiv.org | Graph World Models for Constrained Epidemic Policy Planning 8h ago · 3 min read Epidemic policy planning often requires coordination between geographical regions, taking into account mobility-driven spillovers and how to make use of limited resources. Existing methods either lack action-conditioned models of coupled dynamics or cannot guarantee per-period feasibility. We present EpiMind, a graph world model framework for constrained ... Read full story

08:59 <Ada> arXiv.org | LLM4Trust: Exploring the Capabilities of Large Language Models for Trust Evaluation Jie Wang, Yanbo Sun, Zheng Yan, Jiahe Lan, Elisa Bertino · 8h ago · 3 min read Trust evaluation plays a critical role in cybersecurity by supporting risk mitigation and decision-making. A variety of trust evaluation methods have been proposed, with learning-based approaches offering high accuracy and automation. However, they often require substantial ground truth, suffer from low training efficiency, lack support for basic trust pr... Read full story

08:59 <Ada> arXiv.org | Benchmarking Attention for Tabular Foundation Models Maximilian Schambach, Clemens Biehl, Sam Thelin · 8h ago · 3 min read Tabular in-context learners such as TabPFN, Mitra, or ConTextTab rely on alternating row and column attention over 2D sequences of latent embeddings. These attention patterns differ markedly from the one-dimensional case in language models: row attention involves longer sequences while column attention operates on much shorter ones, and the strided memory... Read full story

08:59 <Ada> arXiv.org | RLVR$^{2}$: Reinforcement Learning with Verifiable Rubric-based Ranking 8h ago · 3 min read Reinforcement Learning with Verifiable Rewards (RLVR) is expanding from tasks with well-defined correctness signals, such as mathematics and code, toward multifaceted quality requirements specified by multi-dimensional rubrics. Since policy optimization consumes one scalar per rollout, rubric-based pipelines must map multiple criterion scores into a scala... Read full story

#ai el 2026-10-12