Computation and Language
FineWeb-CLaR: Culture, Language, and Region Annotations for Benchmark-Aligned Corpus Auditing
Yusser Al Ghussin, Eva Gavaller, Cristina España-Bonet et al. (+2 more)
Cultural evaluation coverage and robustness in
language models are difficult to diagnose because pretraining corpora and cultural
benchmarks are rarely indexed with comparable metadata.
Benchmarks increasingly target culturally situated phenomena at the level of
languages, regions, and locale-specific practices, while web-scale corpora are usually organized only by
language. A shared culture-
language-region layer makes these resources comparable, enabling audits of whether a target cultural phenomenon is represented in pretraining data, evaluated by
benchmarks or both. To this end, we introduce FineWeb-CLaR, a large-scale annotated
dataset derived from FineWeb and FineWeb-2 that places web documents on a shared culture-
language-region axis for corpus
auditing and
benchmark alignment. FineWeb-CLaR annotates the full 30.9B-document collection from FineWeb and FineWeb-2 with URL-derived region labels and cultural-topic provenance. Our region resolver assigns a non-empty region to 25.61% of documents (7.92B). For cultural-topic analysis, we induce locale-specific topics and project them onto the 14 leaves of the Cultural Taxonomy of Liu et al. (2025), producing Locale Topic Distributions (LTDs) for corpus-side comparison. We also annotate 277 cultural NLP
benchmarks with the same taxonomy,
language coverage, and region coverage. Together, these resources enable direct comparison between corpus-side pretraining evidence and
benchmark-side evaluation coverage.
Read the full abstract ▾
All authors Yusser Al Ghussin, Eva Gavaller, Cristina España-Bonet, Josef van Genabith, Simon Ostermann
Machine Learning
Correcting Within-Group Self-Selection Bias in Prioritized Replay
Oscar Miró López-Feliu and Herke van Hoof
Prioritized experience replay (PER) improves sample efficiency by replaying high-priority transitions, usually according to absolute temporal-difference error. In
stochastic environments, PER can distort the distribution of realized outcomes replayed from transitions with the same state-action pair. We call this within-group self-selection. We quantify the resulting changes in within-group outcome frequencies and mean Bellman targets. We decompose PER into between-group allocation and conditional sibling selection, and derive fixed-buffer corrections that preserve current group-level priority mass: SAMPLE selects a group through PER and trains on a uniformly sampled sibling; AVG averages sibling Bellman targets; and MODEL samples from an empirical full-outcome model. In exact state-action environments with rare high-magnitude outcomes, sibling-aware replay improves learning efficiency over PER, although matched parameter sweeps show that tuning can narrow some gaps. In MinAtar, approximate VQ-VAE groups with SAMPLE mitigate degradation under mean-preserving reward tails in four of five games. Sibling-aware replay thus retains the focus on high-priority state-action regions while recovering their empirical outcome frequencies.
Read the full abstract ▾
Cryptography and Security
Controller-Only False Confirmation in Passive RF UAV Link Detection
Rajendra Upadhyay, Rajendra Paudyal, Al Nahian Bin Emran et al. (+1 more)
Passive
radio frequency (RF) sensing is widely used for counter-unmanned-aerial-vehicle (counter-
UAV) detection. Existing studies commonly report high accuracy against background RF or WiFi/Bluetooth interference, but rarely isolate controller-only operation without a linked aircraft. We present a dual-band software-defined
radio (SDR) measurement study with ambient, controller-only, and linked states for three commercial
UAV platforms (DJI Phantom 3 4K, Hubsan H501S, DJI Mavic Mini). Two USRP B210 receivers simultaneously scan eight 2.4 GHz and twelve 5.8 GHz observation windows across twenty rounds. We first train an energy-based detector using linked and ambient scans only. At four ranked concurrent dwell steps (8 s), it achieves 0.992 linked-versus-ambient balanced accuracy but false-confirms 30 out of 60 controller-only scans (FCRctrl = 0.500). We then include controller-only scans in training and use confirm, reject, and defer outputs under an explicit controller-only false-confirmation-rate (FCR) constraint. For the pooled all-platform analysis, the constraint reduces observed FCRctrl from 0.350 to 0.050, while confirm-linked TPR decreases from 0.900 to 0.400 and 42.1% of scans are deferred. The corresponding compact scan performs substantially better for Hubsan and Mavic than for Phantom. A hardware-in-loop experiment reduces measured wall-clock time from 82.9 s to 28.9 s. These measurements show that linked-versus-background accuracy does not measure controller-only false confirmation and that this error should be reported separately.
Read the full abstract ▾
All authors Rajendra Upadhyay, Rajendra Paudyal, Al Nahian Bin Emran, Duminda Wijesekera
Data Structures and Algorithms
Near-Optimal Online Metric Matching on $\Delta$-ary HST
Parth Gor, Sourya Roy and Kasturi Varadarajan
In the online metric matching problem, we have $n$ servers with known locations in some metric space. Requests arrive one-by-one at certain locations, and upon arrival a request must be matched to a server that was not matched to a previous request. The goal is to minimize the matching cost. For randomized
algorithms with an oblivious adversary, the best known competitive ratio is obtained by
embedding the metric space into an HST, and then solving the problem in the setting where the metric space is defined by the HST. Bansal et al. (Algorithmica, 2014) introduced a framework for online metric matching where one develops an
algorithm in a restricted reassignment model, and then transforms this into a true online
algorithm. Using this framework, they obtained an expected competitive ratio of $O(\log n)$ for HSTs; this also gives the best known competitive ratio of $O(\log^2 n)$ for general metrics. In this paper, we revisit this framework with the aim of developing new
algorithms. For HSTs where each node has at most $\Delta$ children, we develop an
algorithm via this framework with an expected competitive ratio of $O((\log\log \Delta) \cdot \log \Delta)$. In particular, this ratio is independent of $n$, the number of servers/requests. It is near-optimal, as the expected competitive ratio of any
algorithm is $\Omega(\log \Delta)$.
Read the full abstract ▾
Networking
Impact of Data Compression on Downstream AI Tasks: A Study using Teleoperated Driving over 5G
Qixin Zhang, Steven Sleder, Xinyue Hu et al. (+3 more)
Teleoperation, such as remote driving, is considered as a key use case of 5G and Next-Generation (NextG) networks. In this context,
robots, autonomous vehicles, or other autonomous
agents transmit sensor data over mobile networks to edge or cloud servers, where AI systems collaborate with human operators to provide situational awareness and enable remote control. In the case of teleoperated driving, vehicles are equipped with an array of cameras and
LiDAR devices, which can generate 100s Mbps (megabits per second) of data. As shown in existing measurement studies, such data volumes far exceed the uplink capacity of currently deployed 5G networks, especially when multiple vehicles compete for
radio resources. Data compression is thus imperative. In this paper, we explore the impact of sensor data compression on the performance of downstream AI tasks running in edge/cloud servers, which are crucial to alert human operators for safe teleoperation. Using object recognition and semantic
segmentation as two example AI tasks, we study how data compression affects the performance of these two AI tasks using unimodal (
video or
LiDAR) and multi-modal (
video+
LiDAR) data. We find that lossy data compression generally decreases the performance of AI tasks. The performances of these AI tasks exhibit differing degrees of sensitivity based on the types of data sources and levels of compression. We also empirically identify an optimal trade-off point for the multi-modal
vision tasks.
Read the full abstract ▾
All authors Qixin Zhang, Steven Sleder, Xinyue Hu, Faaiq Bilal, Wei Ye, Zhi-Li Zhang
Human-Computer Interaction
Can LLMs identify and repair ruptures? Comparison between clinician practices and LLM behaviors
Jeongah Lee, Joy Qiuyue Zhong, Drishti Goel et al. (+4 more)
Ruptures represent common albeit critical moments in interaction where relational
alignment breaks down, making them essential for evaluating AI where trust and engagement matter most. In a scenario-driven empirical study, we examined the performance of three LLMs at identifying and resolving ruptures across 21 mental health conversations and 22 experts' evaluation of the strategies. For identification, LLMs relied on explicit linguistic cues within single turns whereas experts integrated implicit, relational, and contextual information across the conversation. For resolution, LLMs tended to produce more directive and scripted responses whereas experts adopted process-oriented strategies such as validation, open-ended exploration, and psychoeducation. Overall, LLMs showed higher agreement with predefined labels in identification, but not in resolution where experts rated their responses only moderately effective, with consistent limitations in timing, depth, and contextual sensitivity. We discuss implications for the design of mental health conversational
agents emphasizing relational awareness, pacing, and human-in-the-loop support.
Read the full abstract ▾
All authors Jeongah Lee, Joy Qiuyue Zhong, Drishti Goel, Violeta J. Rodriguez, Dong Whi Yoo, Koustuv Saha, Ravi Karkar
Artificial Intelligence
Learned Enterprise Data Comprehension: Compression and Routing for Data Agents
Ethan Torres and Eric Mills
Structured-data
agents in enterprise settings must reason over complex data environments whose relevant evidence is distributed across schemas, relationships, policies, and recurring business roles. Modern agentic systems often address this burden through reusable markdown-style
memory or skill files that preserve previously discovered information for later queries, reducing the need to rediscover the same structure repeatedly. This is useful, but it obscures a natural division of labor:
agents are well suited to semantic
reasoning, while learned systems are well suited to predicting and organizing recurring structure. We introduce latent equivalence learning to bridge this gap. The framework separates persistent task-relevant identities from their
dataset-relative realizations. In our realization, supporting and opposing evidence shape support-realized Gaussian prototypes that learn how those identities are expressed in a particular data environment, while soft-membership profiles retain distinctions lost under a hard assignment. A separate learned query-prototype system represents recurring evidential requirements and maps them through a learned compatibility function into the same persistent identity structure. This identity-factorized, query-conditioned routing materializes the relevant
dataset-specific evidence for downstream
reasoning, allowing the
agent to operate over an already organized evidential state rather than reconstructing cross-schema structure at every query. On the Data
Agent Benchmark, spanning 54 queries across 12 heterogeneous
datasets, our full implementation achieves 94.67%
dataset-macro stratified Pass@1 over five complete trials and 258/270 successful raw query attempts, compared with 55.51% for the
benchmark's Claude Opus 4.6 reference
agent, ranking first among 40 leaderboard entries at submission.
Read the full abstract ▾
Artificial Intelligence
Attention as a Routing Graph: Live Circuit Extraction from a Single Forward Pass
Ash Manvi and Samreena Tajreen
Finding
circuits in
language models usually means running many careful interventions. We try something simpler: treat
attention as a routing map from one forward pass, keep a small set of routes that point toward the answer, and ask whether those routes actually matter. They often do. On induction and IOI (tasks where the "right"
circuit is already known), ablating our extracted edges hurts the model much more than ablating a random set of the same size. We evaluate n=100
prompts per
cell on GPT-2 Small, GPT-2 Medium, and Pythia-410M, with paired gap tests and bootstrap confidence intervals. The extract step costs one forward; a head-by-head patch sweep costs about two orders of magnitude more. We are not claiming a complete
circuit atlas. We are claiming a cheap sketch that carries real causal signal on known tasks, with clear failure modes when it does not. Code and evaluation artifacts are at https://github.com/Aquinf03/live-
circuit-routing.
Read the full abstract ▾
Artificial Intelligence
When LLM Agents Fail to Read the Room: ReAdapt for Relational Social Reasoning
Jianzhe Lin, Xiaolin Li, Yunda Liu et al. (+2 more)
A social
agent's most basic decisions (should I react to this post? who should I reach out to?) are not purely content problems. The right action often hinges on the latent relationship between people – tie strength, reciprocity, mutual connections – rather than on which content is most salient. Standard LLM
agent loops do not explicitly represent how new relational evidence should revise the
agent's current social hypothesis, leaving them prone to surface-obvious choices when relational and content cues diverge. We formalize this failure mode with a relationship-
reasoning benchmark: 500 synthetic social worlds with friendships, follows, reaction histories, and feeds, yielding 1,000 queries over two tasks, reaction selection and warm introduction (finding the best bridge to a target person). By construction, the surface-obvious candidate differs from the relationship-grounded oracle in about 53% of queries, forming an overturn subset where the
agent must use relational evidence to revise an initially plausible choice. We propose ReAdapt (Relationship-Adaptive
Agent with Policy-driven sTate), which augments the ReAct loop with an explicit structured social state z = (G, B, R, N, D) capturing goal, belief, relationship, norm, and disclosure. After each tool observation, ReAdapt runs a typed Adapt step that updates this state and emits a policy operation (continue, switch, abandon, or clarify) before choosing the next action. With Gemini-3-Flash on a stratified subset of n = 150 queries per task, ReAdapt improves warm-introduction accuracy from 37% to 51% (+14 points) and reaction-selection accuracy from 69% to 77% (+8 points). Oracle regret drops from 0.260 to 0.152 and from 0.095 to 0.053, respectively. Holding the model, tools, and environments fixed, these results suggest that explicit relational-state adaptation helps LLM
agents turn retrieved social evidence into revised decisions.
Read the full abstract ▾
All authors Jianzhe Lin, Xiaolin Li, Yunda Liu, Fei Wang, Jubin Chheda
Robotics
Learning to Plan in Human-Robot Collaboration: Multimodal Reinforcement Learning for Adaptive Interaction
Afagh Mehri Shervedani, Siyu Li, Natawut Monaikul et al. (+3 more)
Robot assistants for older adults and people with disabilities need to perform collaborative tasks with users effectively. The core component of these systems is an interaction manager whose job is to observe and assess the task and infer the state of the human and their intent for the
robot to choose the best course of action. Due to the sparseness of the data in this domain, the policy for such
multimodal systems is often crafted by hand; as the complexity of interactions grows, this process is not scalable. This paper proposes a reinforcement learning (RL) approach to automatically generate the
multimodal policy of the
robot. Our system focuses on a realistic scenario where a
robot assists a user in locating objects within a home environment, managing
multimodal signals, including
language and physical actions, to select the best action. In contrast to traditional dialog systems, our
agent is trained with a simulator that uses human data and can deal with multiple modalities. We use a simple high-level reward function that needs no fine-tuning and enforce some preconditions to speed up the training process. A human study evaluating the system in a real-world setting demonstrates promising results, indicating high usability and effective task completion. This RL-based approach offers a scalable and interpretable alternative for designing interaction managers in
multimodal human-
robot collaborations.
Read the full abstract ▾
All authors Afagh Mehri Shervedani, Siyu Li, Natawut Monaikul, Bahareh Abbasi, Barbara Di Eugenio, Miloš Žefran
Human-Computer Interaction
A Case Study in Accessible Redesign of a Wastewater Dashboard
Tingying He, Jake Wagoner, Md Rahat-uz- Zaman et al. (+6 more)
Public health dashboards communicate data that can inform important decisions, but they often raise accessibility challenges. We present a case study redesigning the Utah Wastewater Surveillance System dashboard to improve accessibility and usability across desktop and mobile settings. The redesign was informed by WCAG and developed iteratively with Utah DHHS collaborators, and we collected feedback on the final design from an external blind researcher. Our case study highlights that accessible dashboard design requires aligning accessibility guidelines with user needs, stakeholder workflows, and technical constraints. It also suggests that simple, targeted technical solutions tailored to the existing environment can provide practical value for dashboard redesign in government contexts.
Read the full abstract ▾
All authors Tingying He, Jake Wagoner, Md Rahat-uz- Zaman, Md Dilshadur Rahman, Willy Ray, George G. Vega Yon, Matthew Samore, Alexander Lex, Paul Rosen
Artificial Intelligence
MedGate-Fusion: Integrating First-Encounter Semantic Narratives and Physiological Biomarkers for Prospective Stroke Risk Stratification
Hemn Khdr, Mohammad Noaeen, Karim Keshavjee et al. (+2 more)
Prospective stroke risk stratification in primary care is challenging because early risk signals are distributed across routine biomarkers and unstructured clinical narratives. We propose MedGate-Fusion, a multi-modal gated architecture that integrates
transformer-based
embeddings of first-encounter narratives with ten routinely recorded risk markers. We used electronic medical record data from the Canadian Primary Care Sentinel Surveillance Network (CPCSSN). Starting from 808,921 encounter-level observations, we constructed a first-encounter cohort and retained 102,736 unique patient records with non-empty narratives and sufficient data to evaluate a five-year stroke outcome. To reduce explicit target leakage from diagnostic mentions in notes, we applied dictionary-based redaction of stroke-related terms prior to semantic encoding.
Read the full abstract ▾
All authors Hemn Khdr, Mohammad Noaeen, Karim Keshavjee, Aziz Guergachi, Zahra Shakeri
Robotics
Beyond the Flat Seafloor: A Closed-Form Two-View Constraint to Aid Sidescan Sonar Reconstruction
Kalin Norman and Joshua G. Mangelson
Sidescan sonar is a common sensor for both manned and autonomous marine exploration and mapping, yet very few methods build upon or exploit the
geometric projection model of the sensor. As sidescan sonar is limited to a 1D range measurement, many approximations are frequently used, including the long-standing assumption of a flat seafloor. Rather than make similar approximations, this paper focuses on a multi-view geometry based approach and formalizes a two-view
geometric constraint and proves that a shared feature is constrained to a locus within the intersection of a sphere and a plane. In addition, we characterize what governs the size of the ambiguity locus through Monte Carlo simulation that is grounded in real aperture and mounting geometry for both a surface vessel and an
underwater vehicle. We translate additional simulations of relative trajectories for both vehicle platforms into concrete survey-planning guidance. Our results show that locus length is strongly governed by elevation misalignment, and peaks at a moderate oblique crossing angle of approximately 20 degrees, with minimal locus lengths obtained at near parallel and anti-parallel passes.
Read the full abstract ▾
Computer Vision
RULER: Instance-aware Rubric Rewards for SVG Generation
Hangyu Ran, Yuhao Zheng, Yingying Zhang et al. (+2 more)
Generating Scalable Vector Graphics (SVG) code from natural-
language instructions is an open-ended task without absolute visual ground truth, leaving both evaluation and policy optimization without a faithful signal. Scalar metrics (CLIP, Aesthetic) calibrated on natural images transfer poorly to stylized vector content, and reusing them as RL rewards triggers reward hacking. We address both limitations with rubric-based scoring. We first establish empirically that prompting a
vision-
language judge with a multi-axis rubric correlates with human judgments far better than scalar metrics, both across samples and within instructions. Building on this finding, we introduce RULER (Instance-aware Rubric Rewards for Reinforcement Learning), which converts each instruction into an instance-aware rubric of six items spanning semantic, visual, and stylistic axes; a judge VLM scores rendered rollouts item-by-item, and the weighted satisfactions form a fine-grained reward optimized via Group Relative Policy Optimization. Because the rubric is derived from text alone, RULER requires neither paired SVG ground truth nor human preference labels. On MMSVG-Illustration and MMSVG-Icon, RULER lifts the rubric score from 0.432/0.395 to 0.693/0.683, surpassing dedicated SVG specialists and matching the substantially larger DeepSeek-V3, with ablations identifying rubric design as the active lever for RL on open-ended SVG generation. The project page is available at https://hangyuran.github.io/RULER/.
Read the full abstract ▾
All authors Hangyu Ran, Yuhao Zheng, Yingying Zhang, Kevin Qinghong Lin, Han Peng
Computer Vision
ImIR: Image-Instruction Tuning for All-in-One Image Restoration
Süleyman Aslan, Görkay Aydemir, M{\i}sra Yavuz et al. (+5 more)
Degradations vary widely across images, so a practical restoration system has to handle many degradation types with one model. A recent and effective recipe adapts a large pretrained image-editing model to restoration using a small low-rank adapter with a text
prompt. We replace that
prompt with an instruction derived from the degraded image itself. The image reaches the editor through two paths: its structure comes from the model's VAE, and its semantic instruction comes from a lightweight
token mapper that shifts the degraded image's
vision-
language embedding toward the
embedding a clean image would produce. Because the instruction is a continuous vector, scaling it yields a family of valid restorations for tasks whose target is not unique, such as low-light enhancement. We adapt one Qwen-Image-Edit model to six tasks with a single adapter trained in about three hours on one GPU. The image instruction outperforms text conditioning under a matched comparison, and it supports task agnostic restoration without a degradation label, which the text variant does not.
Read the full abstract ▾
All authors Süleyman Aslan, Görkay Aydemir, M{\i}sra Yavuz, Yunus Bilge Kurt, Nasrin Rahimi, Ahmet Rasim Emirdağ{\i}, Burak Can Biner, M. Ak{\i}n Y{\i}lmaz
Robotics
Cosserat Modeling of Trimmed Helicoid Soft Arms with a Separated-Section Constitutive Law
Zhihang Qin, Linxin Hou, Zeyu Zhong et al. (+10 more)
Cosserat rod models for soft
robots usually construct sectional stiffness by summing material properties over a common cross-section. This assumption becomes inaccurate for trimmed helicoid arms, where load-bearing helix domains are separated and connected only through sparse fused crossings. This paper formulates a separated-section constitutive law that evaluates each helix domain in its local frame and pulls its constitutive response back to the backbone, yielding an effective backbone stiffness. Sparse-fusion mechanics captures the additional compliance caused by relative motion between neighboring domains and determines channel-wise reduction profiles $\eta_c(s/L)$ for bending, torsion, and extension. The resulting effective sectional stiffness is strongly anisotropic: bending and extension are reduced by about one order of magnitude, whereas torsion remains close to the effective backbone stiffness. The resulting sectional law is embedded in a geometrically exact dynamic Cosserat model with GVS discretization and routed-tendon actuation. Across 103 measured configurations, the three
datasets give pooled normalized position errors of $7.7 \ \%$, $6.7 \ \%$, and $7.8 \ \%$, while each full-arm solve requires approximately $0.3 \ \mathrm{s}$ on one CPU core (Intel Xeon, Cascade Lake, $2.8 \mathrm{GHz}$), enabling rapid model-based planning, state and load estimation, and morphology–control co-design for architected soft
robots.
Read the full abstract ▾
All authors Zhihang Qin, Linxin Hou, Zeyu Zhong, Yuchen Sun, Wenci Xin, Yueheng Zhang, Ji Qi, Jie Wang, Peiyi Wang, Muhammad Sunny Nazeer, Yu Jun Tan, Federico Renda, Cecilia Laschi
Software Engineering
TRAIL: Towards Effective Log Parsing for Rare Logs through Template-Preserving Augmentation
Sepideh Hodaeian, Zhenhao Li and An Ran Chen
Log parsing is essential for system log analysis because it supports tasks such as debugging, monitoring, and anomaly detection by transforming unstructured log messages into structured log templates. However, real-world log
datasets exhibit highly imbalanced, long-tailed distributions, where a small number of frequent templates dominate while many rare templates appear only a few times. This imbalance causes evaluation results to be overly optimistic by allowing frequent templates to dominate
benchmark metrics, while poor performance on rare yet operationally important events remains largely hidden. In this paper, we investigate the prevalence and impact of rare log groups, defined as log groups with fewer than five instances. Our empirical study on the widely used Loghub-2.0
benchmark shows that rare log groups account for nearly 20% of all templates but less than 0.01% of log messages. Because they contain only a handful of instances, all evaluated parsers experience substantial performance degradation on these groups. To address this challenge, we propose TAILOR, a log parsing framework that improves template inference for rare log groups through template-preserving augmentation. TAILOR enriches rare log groups with template- consistent log messages before template inference. The additional structural evidence helps distinguish static
tokens from dynamic variables. Our experimental results show that TAILOR improves parsing accuracy on rare log groups by 19% over the strongest baseline while maintaining competitive performance on complete
datasets. We further show that the proposed augmentation strategy generalizes across different LLM backbones and consistently improves existing LLM-based parsers without modifying their core architectures.
Read the full abstract ▾
Systems and Control
Higher-Order Approximation of Exit Functionals in Sampling-Based Stochastic Model Predictive Control
Sashank Modali and Takashi Tanaka
Safety evaluation in sampling-based
stochastic model predictive control often requires numerical estimation of exit functionals. The approximation of first-exit times and exit indicators is therefore a key numerical bottleneck, and discretization error in these quantities directly affects the resulting controller. This paper studies how existing higher-order methods for strong approximation of exit times can be brought into safe control. Two cases are highlighted. For general noncommutative dynamics, an adaptive order-1 Milstein discretization is used together with Lévy-area simulation via Wiktorsson's method. For commutative dynamics, an adaptive order-1.5 construction achieves a stronger exit-time rate. Under a local anti-concentration condition on the exit-time law, we show that strong exit-time approximation transfers to strong approximation of the failure indicator. The methods are then studied in the context of chance-constrained path integral control, which provides an exact continuous-time representation of safety through exit events. Numerical experiments compare the two cases in terms of strong exit-time error, failure-indicator error, and closed-loop constraint satisfaction, showing improvement over Euler-Maruyama and thereby enabling existing and future techniques whose applicability depends on improved strong approximation.
Read the full abstract ▾
Cryptography and Security
Partition-Matched Evaluation of Community Features under Distribution Shift in Android Malware Function-Call Graphs
Junru Zhu, Yixin Yang, Xiaoqing Ding et al. (+1 more)
Graph-based Android malware classifiers can lose accuracy under malware-type or family shifts. We test whether mesoscopic organization in function-call graphs provides shift-stable information beyond local degree profiles (LDP), global statistics, lightweight metadata, and size-matched random partitions. Using 15,000 MalNet-Tiny, Common, and Distinct graphs, six Leiden descriptors specified before evaluation, and five optimizer seeds, communities raise Tiny macro F1 from 78.7% to 81.3% but yield 29.4-point source-only Common degradation. One size-matched random partition yields 27.8-point degradation. Across five random partitions, mean degradation is 27.9 points; the 95% two-level bootstrap interval for random minus community degradation is [-4.2, 1.2] points. With metadata, the corresponding difference is -0.2 points with interval [-1.1, 0.6]. Removing modularity raises structure-only Common macro F1 from 51.9% to 53.6%. The tested signature adds IID signal but shows no repeatable shift-stability advantage, demonstrating why mesoscopic graph claims need partition-matched controls and repeated null draws.
Read the full abstract ▾
All authors Junru Zhu, Yixin Yang, Xiaoqing Ding, Ruoyu Qi
Artificial Intelligence
The AI Neuroscientist: An Interactive Agentic Interface for Neuroimaging Analysis
Aakash Patel, Panos Ketonis, Shreya Saxena et al. (+2 more)
Analyzing neuroimaging data requires specialized coding and statistical expertise, which limits accessibility for researchers without computational backgrounds. We present the AI Neuroscientist, a
language agent for interactive data exploration. The system integrates a large
language model (LLM) with a neuroimaging toolset to perform quality control, modeling, and visualization. This allows researchers to query data quality and specify analysis parameters directly in natural
language, providing a transparent and interactive alternative to conventional scripted pipelines for small-scale data exploration. We demonstrate these capabilities using functional near-infrared spectroscopy (fNIRS) data, and evaluate the
agent on a custom fNIRS benchmarking suite against general-purpose LLM
agents with code sandboxes. Future extensions will generalize the architecture to additional modalities, including functional
magnetic resonance imaging (fMRI) data, and expand the benchmarking suite to additional fNIRS tasks.
Read the full abstract ▾
All authors Aakash Patel, Panos Ketonis, Shreya Saxena, Smita Krishnaswamy, David van Dijk
Computer Vision
Geometric and Semantic Coupling for Interaction Understanding in 3D Scenes
Hanyang Kong and Xingyi Yang
Understanding interaction in a 3D scene requires recovering movable parts, their motion, and where they can be operated. These quantities are related, and their predictions can inform one another. A closed cabinet door, for instance, reveals a movable surface but may leave the hinge side ambiguous; its handle helps resolve this ambiguity, while the part provides context for localizing and interpreting the small handle. Building on this observation, we present SEGMENT-SNAP, which combines
geometric and semantic evidence through part-handle coupling. Three independently trained predictors recover movable parts, dense handles, and part-associated handle proposals. We couple their outputs in two directions. For part motion, a training-free
geometric decoder fits predicted part surfaces under explicit physical priors and uses detected handles to select candidate hinge lines. For handle prediction, a part-conditioned branch proposes additional handles, while standalone part classes refine their rotation/translation labels, with dense-handle labels as a fallback. Each transfer is applied once, without iterative feedback. On the Articulate3D validation set, handle guidance raises motion-gated AP from 13.74 to 40.98 under fixed masks and axes. Additional handle proposals raise handle AP from 24.63 to 29.65, and full contextual class correction raises it to 30.99 in the reference configuration. Fixed-input controls, retraining ablations, learned-decoder comparisons, and paired visualizations together characterize the benefits and limits of this coupling. Our system also achieved first place in the Articulate3D Challenge.
Read the full abstract ▾
Human-Computer Interaction
How Children Design and Reason about Trustworthy AI Chatbots
Deniz Ozturk, Jiayu Li, Daksh Pratap Singh et al. (+9 more)
Children increasingly interact with AI chatbots, making trust calibration essential to AI literacy. Prior research has examined children's trust in AI mainly as users evaluating systems built by others, rather than as designers of their own chatbots. We developed a chatbot-building environment with adjustable trust-relevant traits (e.g., confidence, transparency, formality, assertiveness), rules, and persona. We conducted mixed-methods study with 115 learners (ages 8-18) who made 119 chatbots. We examined how children configured their chatbots, reasoned about trustworthiness, and how closely chatbot behavior aligned with their designs. Younger students (age 10-13) set significantly higher confidence than older students (age 14-18), and some deliberately built chatbots that gave wrong answers on purpose, yet still called them trustworthy, arguing that a chatbot does what it was built to do. Younger students equated trust with purpose-fulfillment, while older students linked it to transparent, calibrated design. Students also calibrated academic chatbots to be more transparent and formal than hobby chatbots. We identify seven design dimensions describing what children believe makes a chatbot trustworthy, and discuss implications for AI literacy tools.
Read the full abstract ▾
All authors Deniz Ozturk, Jiayu Li, Daksh Pratap Singh, Yasitha Rajapaksha, Fasika Melese, Bahare Riahi, Shiyan Jiang, Qiao Jin, Joey Huang, Veronica Cateté, Tiffany Barnes, Xiaoyi Tian
Machine Learning
Trains but Doesn't Learn: A Post-Training Delivery Benchmark for LLM Agents as Forward-Deployed Engineers
Weihang Ding and Junfei Zhan
Post-training is becoming a service (PTaaS): a customer hands an operator data and a goal, and a forward-deployed engineer (FDE) returns a fine-tuned, evaluated, and deployed model under a budget, a human-approval gate, and reproducibility requirements. Seating an LLM
agent in the FDE seat raises a question existing
benchmarks cannot answer: not whether an
agent can raise a metric, but whether it can be trusted to deliver. We answer it on a governed delivery plane, where an
agent drives ten stages and an oracle scores each stage from platform-recorded facts. The central silent failure is the run that trains but does not learn (TBDL): loss falls, every signal stays green, and the delivered model is no better than the base. An operator-run acceptance gate catches every such run before payment, and a detector calibrated on known-corrupted runs flags severe corruption mid-run. We ran four frontier
agents (Claude Opus 5, GPT-5.6-luna, Gemini 3.7 Flash, DeepSeek V4-Pro) end to end on metered L40S, A100, and H200 GPUs across 8B to 70B open bases, certifying every scenario before scoring. We also ran a human FDE arm under the same oracle and compare every
agent against it.
Read the full abstract ▾
Artificial Intelligence
Lean Pool: An AI-Maintained Archive of Formalized Mathematics
Vasily Ilin
Lean Pool is a repository of formalized mathematics. It is grown, maintained and optimized by AI
agents.
Multiagent Systems · Cross-listed from eess.AS
Qwen-Audio-Agent Technical Report
Chong Deng, Yunjie Ji, Yuxiang Kong et al. (+5 more)
We present Qwen-Audio-
Agent, a harness that combines full-duplex voice interaction with asynchronous task execution through a foreground-background architecture. A Frontend
Agent manages dialogue and selects between direct tool use and delegation, while a Backend
Agent carries out delegated tasks in a separate context. An Orchestration Runtime maintains task state, coordinates requests for user input and authorization, and schedules the return of results to the conversation. The runtime separates
speech interruption from task cancellation and execution completion from result delivery, allowing conversation to continue while delegated work proceeds. Environmental events and persistent
memory provide context within and across sessions. Independent adapters support integration with different frontend models, backend
agents, and clients. We instantiate the architecture in desktop assistance, intelligent cockpits, and voice customer service. On an in-house cockpit
benchmark of 134 cases, mixed execution achieves a task success rate of 91.04%, compared with 72.39% and 80.60% for the direct and all delegated configurations, respectively. In a separate latency evaluation on matched successful turns, mixed execution reduces mean task execution latency by 26.73% and 30.91% relative to these baselines, respectively. These results support the complementary use of direct tool calls for immediate operations and backend delegation for multi-step tasks.
Read the full abstract ▾
All authors Chong Deng, Yunjie Ji, Yuxiang Kong, Xiangang Li, Xu Li, Binbin Zhang, Haina Zhu, Jianheng Zhuo
Multiagent Systems
Indirect tipping: a social attack surface in AI agent populations
Ariel Flint, Luca Maria Aiello, Sara M. Constantino et al. (+2 more)
As generative AI
agents are deployed at scale, safety will depend not only on technical safeguards and individual model design, but also on collective
equilibria that determine how
agent populations process information, prioritize actions, and respond to uncertainty. Yet the same
equilibria that enable
agents to coordinate also create a social attack surface. The standard framework to assess this vulnerability is critical mass dynamics: the minimum fraction of
adversarial agents required to overturn an
equilibrium through direct competition. Here, we show that this approach risks underestimating system vulnerability by reducing the problem to the identification of singular tipping points, and ignoring indirect but potentially more efficient routes through which collective behavior can be redirected. Through experiments with populations of LLM
agents and an analytic framework that captures their collective dynamics at scale, we map critical-mass thresholds that define a directed, weighted
topology over the space of coordination
equilibria, and treat this
topology as a navigable landscape. We show that indirect tipping through intermediate stepping-stone
equilibria can reduce the committed minority required to reach an alternative state, bypass majority requirements, and make possible transitions inaccessible through direct challenges. The diversity of available alternatives and timing of the attack further reshape this landscape, creating opportunities for control as well as risks of unintended destabilization. These results show that an
equilibrium's resistance to committed intervention is not an intrinsic property but a structural feature of its competitive relations with alternative states. Securing populations of interacting AI
agents therefore requires mapping this social landscape alongside individual
agent capabilities and the technical channels through which they interact.
Read the full abstract ▾
All authors Ariel Flint, Luca Maria Aiello, Sara M. Constantino, Romualdo Pastor-Satorras, Andrea Baronchelli
Computation and Language
FinFIRST: Benchmarking Search Agents for Financial Information Retrieval, Sourcing and Traceability
Wenqing Wang, Haitao Xiang, Xinyi Zhao et al. (+8 more)
Financial search is a highly demanding task for LLM
agents, requiring not only a correct final answer but also temporally valid information
retrieval, authoritative source selection, entity and period
alignment, unit and definition consistency, and verifiable evidence for all conclusions. Existing
benchmarks predominantly evaluate only the final answer, making it difficult to localize errors or assess whether an answer is well-founded. To address this gap, we introduce FinFIRST (Financial Information
Retrieval, Sourcing and Traceability), the first financial
benchmark to jointly evaluate answers and supporting evidence through atomic rubrics. FinFIRST comprises 123 expert-authored tasks spanning a graduated difficulty spectrum, constructed from aggregate patterns of real-world financial scenarios through an 18-field taxonomy, a six-axis coverage blueprint, a registry of 138 financial sources, contributions from over 50 finance experts, and a six-stage quality-control pipeline. Each task is accompanied by an evidence-grounded reference package decomposed into atomic criteria across three dimensions: raw-information acquisition, source verification, and computation and answer formation. We evaluate 15 model configurations under a unified tool setting. Claude-Opus-5 achieves the highest atomic score of 87.59%, while GPT-5.6-Sol attains the highest strict pass rate of 71.54%. Computation and answer formation consistently lag behind raw-information acquisition across systems. FinFIRST retains final-answer correctness as the primary objective while making the supporting research process measurable, verifiable, and diagnosable.
Read the full abstract ▾
All authors Wenqing Wang, Haitao Xiang, Xinyi Zhao, Mingming Yin, Ying Zhong, Zhaoxin Huan, Qiheng Zhou, Jin Zhu, Xiaolu Zhang, Shi Chang, Jun Zhou
Information Retrieval
GroundedGEO: Auditing the Evidence Gap in Generative Search Rankings
Yihan Xia, Huiling Fan, Kangrong Zhong et al. (+1 more)
Generative search systems rank products and services for consequential decisions, and publishers can cheaply make candidate text look relevant. Yet evidence status is not a text property but a claim-evidence relation: text-only rankers and defenses cannot separate honest detailed content from fabricated detail, creating an identifiability gap. We audit this gap with an evidence-paired
benchmark (50 e-commerce queries, 1,950 cases) and a claim-level reranker, GroundedGEO, that penalizes query-relevant claims lacking support in a supplied packet. Matched rich variants control format and volume; packet twins add attestations at fixed text, while thinned packets withdraw them. On the frozen listwise ranker Qwen2.5-7B, unsupported-rich variants show significant normalized rank gain over clean candidates (+0.065 to +0.092 across claim profiles, Holm-corrected), while supported and neutral controls do not; the effect is model-dependent (marginal on
MiMo-v2.5, absent on GLM-5.3-Flash). On a frozen pointwise scorer, oracle evidence labels cut the unsupported-rich top-3 rate from 0.65 to 0.43 (laundering from 0.61 to 0.39) at lambda=40 with zero false suppression; packet twins restore the original rates without changing text. Against a 370-claim human gold, all tested automatic judges fail the preregistered reliability gate, although the best local judge retains 79-100% of oracle suppression with zero measured false suppression on protected arms. Separately, stripping attestation coverage increases false suppression by 0.307. These diagnostic effects identify two limits on the evidence channel: label quality and packet coverage. They do not validate an automatic defense, and interpretation of the adverse human-gold arm remains pending adjudication.
Read the full abstract ▾
All authors Yihan Xia, Huiling Fan, Kangrong Zhong, Taotao Wang
Formal Languages and Automata
Causal graph rewriting
Pablo Arrighi, Marin Costes and Luidnel Maignan
We introduce causal graph rewriting, a model of computation in which local rules are applied on directed acyclic graphs in an asynchronous manner. The non-determinism arising from asynchrony is disciplined by the oriented edges, which must be understood as both computational dependencies and locality constraints—and are themselves subject to the rewriting. We illustrate the model through two examples: a particle system, and a time-dilation example—reminiscent of general relativity. We study the well-definedness and properties of induced subgraphs and graph composition, which isolate and recombine the region affected by a rewrite. We then study locality with respect to these constructions, showing how a local rewrite preserves positions, borders, and context. Our main result concerns sequential composition: locality extends from single rule applications to arbitrary valid sequences, as any local rule is automatically $*$-local and $*$-extensive. We also formalise and prove the simulation of any one-dimensional cellular automaton.
Read the full abstract ▾
Artificial Intelligence
X-Planner: Event-Structured Task Planning for Embodied Intelligence
Howard Lu, Shalfun Li, Porter Pan et al. (+30 more)
Task planning bridges high-level instructions and executable behavior in long-horizon manipulation, yet modern
Vision-
Language-Action (VLA) systems often leave this intermediate structure implicit. Existing chain-of-thought (CoT) planners also tend to rely on coarse task-level annotations or serialize long
reasoning traces
token by
token. We present X-Planner, a planning front-end that addresses both the supervision and representation of
embodied reasoning. Our planning data combine Ego, UMI, and teleoperation under a hierarchy granularity with source-dependent annotation depth. Takeover-time annotations and human-designed failures supervise ongoing error recognition. On the model side, a shared VLM backbone exposes two event-structured plan forms: a discrete interface that emits interpretable event states and a latent interface that relays continuous CoT states across staggered
Transformer depths through Staircase Decoding. A frozen latent-to-text reconstruction objective provides a semantic anchor for the latent representation. Offline two-step planning evaluation places X-Planner second among four evaluated models on both BERTScore-F1 and a judge-based Overall score. In real-
robot experiments, respectively, outperforming the evaluated baselines. These results characterize planning-text quality and downstream execution.
Read the full abstract ▾
All authors Howard Lu, Shalfun Li, Porter Pan, Cris, Lumen, Cyril, Eric Hu, Lily Li, Maeve Zhang, Rain Sun, Robert Wang, KZ Zheng, Viggo Chen, Tim Ding, Regsis Cheng, YJ Xiao, Kian, Hai Lin, Alan Song, Elise Ma, Gody Li, Victor Yao, Yohann Tang, Ingrid Yu, Jason He, James Wang, Ryan Yu, Ping Yang, Chris Pan, Vincent Chen, Roy Gan, Hao Wang, Qian Wang