arXiv subjects

Search papers

Search indexed arXiv papers on artificial intelligence, large language models, computer vision and robotics. Read source abstracts and follow links to arXiv.

At least 901 records · Page 50Linked to original sources

arXiv2026

Micro-Doppler Sensing for UAV Flight-State Estimation in ISAC Networks

Integrated sensing and communication (ISAC) provides a promising foundation for low-altitude unmanned aerial vehicle (UAV) monitoring. However, most existing studies focus primarily on position and bulk velocity, thus providing limited information about UAV rotor dynamics and orientation. To address this limitation, this paper develops an orthogonal frequency-division multiplexing (OFDM)-ISAC framework that exploits UAV micro-Doppler signatures to estimate operational states, including rotor rate and rotor-axis orientation, thereby providing richer UAV state information beyond conventional ISAC localization and velocity sensing. The framework progresses from single-station micro-Doppler sensing to multi-station rotor-axis estimation. At each base station, micro-Doppler signatures are first extracted and used to estimate the rotor rate. The framework is then extended to a distributed ISAC network, where view-dependent micro-Doppler signatures from multiple sensing directions are fused to estimate the common rotor-axis direction. Since the fusion relies on Doppler-bandwidth features rather than complex channel phases, inter-station carrier-phase synchronization is not required. We show that three geometrically diverse views can locally identify both the rotor-axis direction and the unknown rotor-tip-speed scale. In addition, a feature-aware radar equation bridges blade radar cross section (RCS), sensing range, transmit power, and observation duration to micro-Doppler detection and rotor-rate estimation accuracy. The effects of residual motion and range migration on the coherent processing interval (CPI) are also analyzed. Simulations and experimental measurements validate the proposed framework and demonstrate the feasibility of rotor-rate estimation and multi-station rotor-axis sensing.

eess.SP↗

arXiv2026

Disk tilts in Milky Way-like galaxies and their link to mergers

We investigated disk reorientation and its relationship to mergers in a sample of 93 Milky Way-mass galaxies drawn from the TNG50 simulations. From this sample, we identified 117 individual disk-tilt events with clearly defined start and end times, and we divided them into two classes based on whether the disk evolves smoothly or undergoes rapid disruption and regrowth during the tilt duration. For each tilt event, we linked the most massive merger that occurs in close temporal association with it. We focus on those tilt events occurring during phases of smooth disk evolution, which typically show tilt rates below 10 degrees per Gyr, persist for 2-9 Gyr, and are generally connected to minor mergers. Roughly 40% of these events reach tilt angles exceeding 30 degrees, most often triggered by satellites whose masses are comparable to or less massive than the galaxy's total stellar mass, merging on tangential orbits with orbital angular momentum that is retrograde or nearly perpendicular to the disk spin. The satellite consistently drives the disk tilt toward alignment with the satellite's orbital spin; satellites that came in on retrograde orbits relative to the disk spin can thus induce large tilt angles and can produce an almost orthogonal configuration between the stellar disk and the satellite's orbital plane as an intermediate phase. Owing to the comparatively slow tilt rates, the stellar disk can keep stable and continue to grow throughout the reorientation process. Consequently, galaxies that experience strong disk tilts can still maintain stellar disks that are as thin as those in systems without significant reorientation.

astro-ph.GA↗

arXiv2026

KBS: A Korean Benchmark Suite for Foundation Model Evaluation

We present a Korean Benchmark Suite (KBS) comprising Ko-ARC, Ko-GSM8K, Ko-EQ-Bench, Ko-WinoGrande, Ko-LAMBADA, and Ko-IFEval, covering six capabilities across 9,396 items. The suite combines review, localization, direct Korean construction, and redesign for Korean, avoiding a translation-only approach to benchmark adaptation. We evaluate Korean-developed models, multilingual open models, and closed API models under protocols that account for model access, prompt format, and scoring differences. Results show that Korean benchmark scores depend on task construction, scoring rules, prompting, and model post-training. These findings suggest that Korean evaluation should report both final scores and the settings used to obtain them.

cs.CL↗

arXiv2026

What Should We Remove? Evaluating the Reliability and Costs of RLVR Data Decontamination

Reinforcement learning with verifiable rewards (RLVR) math corpora are routinely screened for benchmark overlap, but the reliability and data-loss costs of cleaning policies remain underexamined. We audit sixteen open RLVR math corpora comprising 1,673,514 corpus-specific training items against six benchmarks containing 1,949 questions. We evaluate eight policies based on exact, lexical, and semantic matching using 594 independently and blindly dual-reviewed training-benchmark pairs. Calibration and held-out sets have disjoint benchmark questions and training-example families. A separate review of all 83 candidate matches for 80 sampled training items measures item-level false removal. At cosine threshold 0.80, 52 of 59 sampled removals (88%) were confirmed false removals; this is not a demonstrated population rate. No semantic threshold in our preregistered grid (0.80-0.98) meets the 90% conservative-precision target on calibration. The existing judge-adjudicated policy achieves 98.1% resolved precision and 98.2% candidate-frame recall on held-out corpora. It has zero confirmed false removals among six sampled removals, with one unresolved case; this does not establish zero risk. Within the evaluated candidate frame, this policy offers a more favorable precision-recall balance than the semantic-threshold alternatives tested. Recall does not measure contamination outside the candidate frame. We release reviewed labels, per-policy decisions, and coverage accounting to support evaluation of cleaning reliability and cost rather than overlap counts alone.

cs.CL↗

arXiv2026

Not All Change Is Damage: Update-Matched Unlearning for Large Language Models

Unlearning aims to remove selected knowledge from large language models to mitigate privacy and safety risks while preserving useful capabilities. For models that continue to be fine-tuned after release, this goal is insufficient: they must also resist relearning the deleted knowledge while preserving plasticity, the ability to learn new tasks. Existing defenses train against simulated relearning while trying to preserve useful capabilities, but low relearning alone cannot distinguish selective resistance from a general loss of plasticity. Evaluating preservation after an update requires deciding what the updated model should remain close to. A pre-update reference penalizes useful changes induced by the update simply because they depart from earlier outputs. We therefore formulate the reference-timing hypothesis that, at comparable target suppression, an update-matched reference better preserves plasticity than a pre-update one. To test and address this hypothesis, we propose Update-Matched Unlearning (UMU), which applies the same simulated relearning update to the defended model and to an undefended reference initialized from the same unlearned model, and uses new-task learning to guide model selection. Extensive experiments on TOFU and FaithUn with Llama and Qwen models show that UMU improves both relearning resistance and plasticity across all three base unlearning methods, lowering harmful and benign relearning by 30.4 and 9.9 points on average while raising new-task accuracy by 3.5 points. When the reference does not receive the update, relearning remains similarly low but new-task accuracy falls below the base method, supporting the reference-timing hypothesis and showing that robust unlearning should prevent recovery selectively without suppressing learning itself.

cs.LG↗

arXiv2026

On the Subadditivity of Mutual f-Information

We address an open problem raised by Polyanskiy and Wu in their textbook concerning the subadditivity of mutual $f$-information. We study two notions of generalized mutual information: mutual $f$-information, defined by $\widetilde{I}_f(X;Y)=D_f(p_{XY}\|p_Xp_Y)$, and variational $(G,f)$-information, obtained by minimizing an average of transformed $f$-divergences. For the first notion, we derive a sufficient condition on $g(x)=x^2f''(x)$ that guarantees subadditivity under conditionally independent observations. This condition yields explicit families beyond the previously known examples generated by $x\log x$ and $(x-1)\log x$. Our proof relies on the concavity of a generalized variance functional. For the variational notion, we extend existing sufficient conditions beyond the transforms $G(x)\in\{x,\log(1+x),-\log(1-x)\}$. Using a comparison inequality for $G^{-1}(G(A)+G(B))$, we obtain differential criteria for concave and convex transforms. Explicit constructions illustrate the resulting classes of subadditive information measures.

cs.IT↗

arXiv2026

Are High-EW LAEs good analogs for the earliest galaxies? A study on the ionizing conditions and gas properties of LAEs at $2\lesssim z\lesssim7

Lyman-$α$ emitters (LAEs) are frequently proposed as analogs of the low-mass, metal-poor galaxies thought to dominate cosmic reionization. In particular, galaxies with large Ly$α$ equivalent widths (EWs) are expected to have not experienced significant chemical enrichment. We test this hypothesis using a sample of 88 LAEs at $2 \lesssim z\lesssim 7$, selected from MUSE observations that are undetected in deep HST imaging and subsequently followed up with JWST/NIRSpec spectroscopy. Despite the unprecedented sensitivity of JWST, only 50 of the LAEs exhibit secure counterparts in available deep NIRCam imaging reaching typical 3$σ$ depths of $m_\text{AB}\sim30.3$ mag, highlighting the intrinsically faint nature of the population. Gas-phase metallicities are inferred by fitting optical emission lines with a grid of BPASS+cloudy photoionization models, while stellar masses are derived from NIRCam photometry through SED fitting. The final sample with secure stellar masses and gas-phase metallicities encompasses only 16 LAEs, occupying regions of standard emission-line diagnostic diagrams consistent with low-metallicity star-forming systems. Although the strong-line metallicity diagnostic R3 appears to decrease with increasing EW, we find no compelling evidence for a strong monotonic relation between EW and gas-phase metallicity when metallicities are inferred through full emission-line fitting. However, galaxies with stronger Ly$α$ emission tend to lie below the average mass-metallicity relation at fixed stellar masses, suggesting that they are preferentially less chemically evolved than typical galaxies of the same mass. We conclude that high-EW LAEs provide useful, though imperfect, analogs of the galaxies thought to drive cosmic reionization, and combining EW with stellar mass may provide a more effective route for identifying the most chemically primitive galaxies.

astro-ph.GA↗

arXiv2026

"Do You Trust It... or Check It?": Exploring Children's Verification Behaviors in Game-Based Human-AI Collaboration

Children are increasingly interacting with AI agents in learning and problem-solving contexts, where AI suggestions may appear fluent, helpful, and socially engaging. However, children may not always be able to judge whether such suggestions are accurate, incomplete, or misleading. This paper presents a game-based human-AI collaboration probe designed to examine how children verify AI suggestions during task-based interaction. The system embeds AI suggestions within a biological knowledge reasoning game and compares three Agent interaction styles: partner-like, tool-like, and dual-state Agent. Instead of treating trust as a static post-task attitude, this study focuses on children's situated verification behaviors, including acceptance, evidence checking, questioning, revision, and continued AI use. A preliminary evaluation with children aged 8-10 combined interaction logs, post-task ratings, and semi-structured interviews. The findings suggest that Agent style shaped children's engagement and response pathways: partner-like interaction tended to support direct questioning, while tool-like interaction was more associated with evidence seeking. Children generally viewed AI as useful but fallible, and their questioning behaviors varied in reasoning depth. The study contributes design insights for supporting children's critical, reflective, and evidence-based use of AI in collaborative tasks.

cs.HC↗

arXiv2026

Compatibility Matters: A Framework for Token-Level Teacher Selection in On-Policy Distillation

On-policy distillation (OPD) has recently been introduced as a promising paradigm for continual learning, where a student policy can be supervised at the token-level. Typically, OPD is performed using a single teacher per prompt rollout (or even per training run). However, this ignores a whole other axis of flexibility that is available due to the on-policy nature - token-level teacher selection during rollouts. Building on this insight, we theoretically explore what makes a teacher and a student compatible for OPD, and leverage our findings to derive several empirical heuristics for teacher selection. In particular, our framework captures several desiderata for OPD: compatible teacher-student reasoning patterns, teacher competence, and induced student diversity. Theoretically, we categorize several types of teacher routing schemes, and we prove that under different scenarios, different routing schemes are likely to be beneficial for optimization and student diversity. We back up our findings with extensive experimental results in domains such as math, coding, kernel generation, and visual reasoning, where we find upwards of 10 percentage points of improvement on pass@$k$ when compared to vanilla OPD and the teacher experts on certain tasks, indicating both capability improvement and diversity over high-quality responses.

cs.LG↗

arXiv2026

Model-independent parameters from DESI DR1 LRG2 with an explicitly varying power-spectrum shape

Distinguishing between dark energy and modified gravity as the origin of the accelerated expansion requires the cosmic distances and the growth of structure to be measured in a model-independent way. Such measurements are difficult to perform while the primordial shape parameters are varied at the same time, and those parameters are therefore usually held fixed at the values determined by cosmic microwave background experiments. In this work, we present a methodology that constrains the shape parameters simultaneously with the model-independent distance and growth parameters, using an emulator built upon a hybrid model in which perturbative predictions are supplemented by templates measured from $N$-body simulations. {Because these templates are measured rather than expanded, the model carries non-linear corrections beyond the order at which its perturbative part is truncated, so that its accuracy on small scales is set by the simulations rather than by the reach of the expansion.} A Gaussian process emulator supplies these templates rapidly across the space of scale-dependent parameters, which makes the joint analysis computationally feasible, and the emulator is trained and validated over a range spanning $5σ$ of the Planck constraints. Applying the method to the LRG2 sample of DESI DR1, we obtain measurements of the Alcock--Paczynski parameters $α_\perp$ and $α_{||}$ and of the growth rate $fσ_8$. {With the tight Planck prior adopted here the constraints differ from those obtained with the shape parameters held fixed by less than a twentieth of their own uncertainties, so that the choice of template contributes at that level to the error budget of the measurement. Projected onto flat $Λ$CDM after the fit, the three measurements are mutually consistent, with $χ^2=0.7$ for two degrees of freedom, and require $h=0.724\pm0.017$.

astro-ph.CO↗

arXiv2026

Workplace Ambiguity and Technology Support for Autistic and IDD Workers

Workplace ambiguity, where expectations, evaluations, or procedures are left unstated, acts as a structural access barrier for autistic and IDD workers, routinely forcing them to act on guesses. To investigate how these barriers operate, we interviewed 11 caregivers, concluding with an AI co-design activity to explore technological solutions. Their perspectives revealed five themes of workplace ambiguity, which we synthesize into a conceptual framework comprising three intersecting dimensions: evaluative, procedural, and relational. To address these barriers, participants proposed three forms of technology support, including ambient real-time redirection, feedback processing, and physiological guidance, designed to make hidden expectations perceptible and actionable. However, this exposes a core tension: technologies designed to make the workplace legible can inadvertently make the worker hyper-legible to employers through surveillance. We argue that ambiguity-aware workplaces must make expectations explicit before they carry consequences, and assistive technologies must remain accountable to the worker.

cs.HC↗

arXiv2026

Compositional Automata Learning of Timed Synchronous Systems

Active model learning is a crucial technique for automatically constructing formal models of black-box timed systems. However, existing learning approaches typically construct a single monolithic automaton for a timed system with multiple synchronous components, obscuring its component structure and hindering component-level analysis. We propose a compositional learning approach that constructs a deterministic real-time automaton (RTA) for each component with access only to the target system as a whole. Upon termination, the learned RTAs form a network that exhibits exactly the same timed behaviors as the target system. Within the respective resource limits, our implementation terminates successfully on 213 of the 240 (88.8%) randomly generated benchmark instances and all six adapted case-study benchmarks, and requires less learning time than monolithic approaches on some instances.

cs.FL↗

arXiv2026

CLIS-MPPI: Closed-Loop Model Predictive Path Integral Control With Initial-State Sampling Under Execution Delay

Sampling-based motion planning such as Model Predictive Path Integral (MPPI) control has become attractive for robotics due to its ability to handle nonlinear dynamics and non-smooth objectives without gradient information, while leveraging parallel rollout computation for real-time performance. However, in real-world robotic systems, the optimized control input is often applied after a non-negligible planning delay, causing a mismatch between the state used in forward simulation and the actual state at control execution, which can significantly degrade planning accuracy and may even lead to unsafe behavior such as collisions. This paper proposes Closed-Loop MPPI with Initial-State Sampling (CLIS-MPPI), which explicitly addresses execution-time state mismatch by modeling planning delay as a distribution over plausible execution-time initial states, performing rollout optimization from sampled initial states, and applying a rollout-based local feedback correction during control execution. The proposed method is validated through real-world experiments on a 6-DoF UR5e manipulator in a narrow-passage task involving obstacle avoidance and end-effector target reaching. Across 5 real-world runs of the same task, standard MPPI collided in 2 runs, whereas CLIS-MPPI remained collision-free in all 5 runs. Averaged over all collision-free runs for each method (3 for MPPI and 5 for CLIS-MPPI), CLIS-MPPI reduced the average task completion time from 25.4s to 24.4s and the average end-effector path length from 0.456 m to 0.425 m. It also reduced the overall joint-position and joint-velocity root mean square errors (RMSEs) from 0.100 rad and 0.108 rad/s to 0.017 rad and 0.011 rad/s, respectively. These results demonstrate improved safety, planning-execution consistency, and motion smoothness.

cs.RO↗

arXiv2026

Representation Redistribution in Sparse Autoencoders

Sparse autoencoders (SAEs) decompose language model activations into sparse features, but prompt transformations designed to preserve the underlying request can redistribute activity across them. A feature change observed on one transformed prompt need not recur, and even a recurring change need not affect a downstream computation. We introduce the redistribution sensitivity score (RSS), which identifies large feature responses that recur in a consistent direction, and evaluate the resulting supports on variants excluded from selection. Across 27 pretrained SAEs spanning eight language models, the tested prompt transformations produce a reproducible redistribution profile on these excluded variants. We then separate recurring representation change from signed readout contribution and decision failure. For a fixed linear readout, an exact decomposition shows how response magnitude, readout coefficients, directional alignment, and the clean margin determine downstream consequences. In two frozen safety readouts, combining RSS support erosion with the full clean margin yields recurrent failure AUROCs of 0.935 and 0.958 using median supports of 20 and 19 features, compared with 0.965 and 0.964 using full erosion. RSS therefore provides a way to identify recurring sparse representation changes and quantify their contribution to a downstream readout without equating representation change with failure. Code is available at https://github.com/OpenSAE/Redistribution-Sensitivity-Score

cs.LG↗

arXiv2026

Range: Decoupling KV-Cache Reuse Granularity from Storage Work

LLM serving reuses cached key-value (KV) state to skip prefill, and growing contexts push that state past GPU HBM and CPU DRAM onto NVMe SSDs, placing prefix restore on the critical path of time-to-first-token (TTFT). We find that on a well-configured host-issued path the SSD is not the bottleneck: restoring the same 6 GiB of KV takes 2.1 s as 192 objects but 3.0 s as 6,144 objects, and more under layerwise loading, because every layer of the storage stack - lookup, allocation, I/O submission, GPU delivery, store, index - charges per logical object while the bytes stay the same. This cost grows with the KV footprint: longer contexts, deeper models under layerwise loading, and tensor parallelism each multiply the object count. Prior systems move this per-object work from the CPU to the GPU; the work remains per object. This paper presents Range, a KV-cache storage engine for vLLM and LMCache built on one principle: every cost on the storage path scales with bytes or physical batches, never with logical objects. Range tickets replace per-key lookup and ownership with token metadata and physical leases; byte stripes make the on-media layout independent of reuse granularity; completion-owned GPU transfer banks batch delivery and extraction and release each layer to attention as it lands. Range keeps the kernel cuFile path and changes four lines of vLLM. On a fixed 6 GiB restore, Range's median TTFT stays at 1.79-1.80 s across a 32x range of object counts and under layerwise delivery, within 70 ms of the prefix already being resident in GPU memory, while tuned in-process LMCache rises from 2.12 s to 3.01 s. Range improves on the fastest multiprocess DRAM-bounced path by 0.44 s at coarse granularity while moving no payload through host memory, and holds layerwise reuse at the cost of coarse-grained reuse, where the best stock path takes 2.24 s.

cs.OS↗

arXiv2026

Enhanced Dissipation under Variable Growth: Stochastic $p(x)$-Laplace Equations with Transport Noise

We study dissipation enhancement for stochastic $p(x)$-Laplace equations with divergence-free transport noise on $\mathbb T^d$, $d\ge2$. Under log-Hölder continuity and degenerate subcritical growth, we establish global variational well-posedness and an exact energy identity. For every prescribed positive coefficient, suitable transport noise yields averaged algebraic decay uniformly on bounded zero-mean $L^2$ balls. The noise is independent of the individual initial datum. This uniform enhancement holds despite the absence of homogeneous scaling and the two flux growth regimes associated with the variable exponent. We also identify $p^+$ as the optimal exponent for uniform modular coercivity near zero and obtain a strict averaged comparison with the unperturbed evolution at any prescribed time $T\ge1$ on initial-data shells with bounded gradient modular. For each prescribed decay rate and finite moment order, deterministic stepwise Fourier profiles with fixed intensity and increasing frequency range give exponential decay of the $L^2$ norm in mean square and almost surely. The absolute random prefactors have uniformly bounded moments of the prescribed order throughout the initial-data ball.

math.PR↗

arXiv2026

A Responsibility-Centered Coordination Mechanism for Decentralized Multi-Agent Systems

Centralized orchestration provides comprehensive multi-agent coordination but creates bottlenecks in scalability, adaptability, and collective performance. Decentralized approaches alleviate these bottlenecks by distributing coordination across agents, yet often leave responsibility for delegated dependencies and failure recovery unclear. To address this gap, we propose RCM, a responsibility-centered coordination mechanism that leverages commitments to couple task delegation with explicit responsibilities of the involved parties. Through recursive delegation, agents retain their upstream responsibilities while locally orchestrating downstream commitments, thereby preserving accountability across delegation chains, creating a network of local orchestration. We operationalize this mechanism locally through responsibility-constrained decision-making and a responsibility-oriented communication protocol (RCP), enabling agents to autonomously establish, manage, and terminate commitments. As agents act in parallel, their local actions continuously evolve the commitment graph, allowing global coordination to emerge and adapt toward a shared mission without centralized control. Experiments across reasoning, coding, and executable agent environments demonstrate that RCM outperforms strong centralized and decentralized coordination baselines by up to 18.7 points in task performance and 10.2 points in Responsibility, while improving failure and harm detection without task-specific training.

cs.MA↗

arXiv2026

System1-Arena: Which Decisions Can Be Left to System 1?

Large language models (LLMs) often make discrete decisions, such as flagging phishing or escalating to a human. System-1 models, such as Jev and the OpenAI Decisions API, return probabilities over options in a single forward pass, while System-2 approaches generate reasoning steps. Which decisions can System-1 models handle, and which require generative reasoning? We introduce System1-Arena, a benchmark that gives all models the same decision requests and scores their probabilities. It contains 92k questions across 105 tasks from 268 public sources, in four groups: text classification and knowledge, applied decisions, evaluation and control, and agent and action. It also includes 138 closed-loop environments, from web shopping and SWE-bench coding control to robotics, games, and driving; in real-time mode, the world evolves while a model decides. We evaluate 30 models: 7 closed LLMs, 13 open LLMs read at the answer token, and 10 System-1 models. The largest System-1 models match or exceed fast closed LLMs in non-agentic accuracy, with better-calibrated probabilities at one-quarter of the latency. A reasoning LLM is the most accurate, at about 70x the cost of a fast LLM. In closed loops, large open LLMs succeed most often (up to 32.9%); small System-1 models rarely succeed.

cs.CL↗