Peer Revenue Publish or profit

Papers

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

Computer Science

cs Watch Computer Science

Wednesday 14 October · still arriving

Artificial Intelligence

Designing Textual Interfaces To Enhance Llm Reasoning On Audio-Visual Question Answering

Yi Wang, Tianzi Wang, Tong Lei et al. (+2 more)

Automatic prompt optimization (APO) has shown strong effectiveness for text-based LLMs and LLM systems, but its application to audio-visual understanding systems requires suitable textual interfaces through which multimodal reasoning can be exposed to text-based optimizers. We investigate how the design of textual interfaces affects prompt optimization for multimodal reasoning, on the audio-visual question answering task. We compare three interface designs, including a direct inference system consisting of a single omni-modal large language model (OmniLLM), a captioner-reasoner system, and an agentic architecture in which a primary agent orchestrates subagent OmniLLMs. Experiments with Qwen3-Omni and Gemini show that direct prompt optimization provides model-dependent gains; the captioner-reasoner system and the agentic system expose more interpretable optimization interfaces but are harder to optimize, where error analysis of cases affected by optimization is conducted. Our study highlights textual interface design as an important factor in making prompt optimization effective for omni-modal reasoning systems. Our implementation details are specified in GitHub repo.
Read the full abstract

All authors Yi Wang, Tianzi Wang, Tong Lei, Dahan Wang, Peter Bell

Human-Computer Interaction

Scoping the Landscape of Large Language Model-based Affective Personal Informatics: Review, Reflection, Future Direction

Minjie Liu, Lihuan Li, Xuran Li et al. (+2 more)

Affective personal informatics (PI) involves capturing, tracking, analyzing, and reflecting on personal data to support emotional self-awareness and mental well-being. Recent advances in large language models (LLMs) are reshaping affective PI, yet a comprehensive survey of LLM-based affective PI systems is lacking. To bridge this gap, we systematically analyzed 45 papers published between January 2020 and June 2026 across four major computing and health databases, examining their macro trends, system taxonomies, and research methodologies. Our synthesis reveals an overreliance on multi-stage, text-centric pipelines, a predominance of active, screen-bound interactions, and a severe demographic skew toward young, university-campus-based cohorts in empirical evaluations. Based on these findings, we critically reflect on prevailing architectural and methodological shortcomings and outline actionable future directions for building end-to-end multimodal, low-burden, and inclusive affective PI systems.
Read the full abstract

All authors Minjie Liu, Lihuan Li, Xuran Li, Flora Salim, Benjamin Tag

Computer Vision

VisualWorldBench: A Fine-Grained Multi-Task Benchmark for Evaluating Visual World Knowledge in MLLMs

Chen Chen, Zheming Liang and Jiaqi Wang

Visual world knowledge is a core capability of multimodal large language models (MLLMs), underpinning knowledge-intensive visual question answering and fine-grained visual reasoning. However, existing benchmarks remain limited in two key respects. First, their category coverage is narrow: most benchmarks span only a small number of categories across a few domains, falling far short of the open-ended scale of real-world visual entities. Second, their task construction is often insufficiently rigorous: models can answer correctly using coarse-grained cues, vague associations, or elimination strategies without truly mastering precise concepts, making it difficult to accurately assess their knowledge of fine-grained entities. We introduce VisualWorldBench, a taxonomy-grounded multi-task benchmark built on a hierarchical taxonomy of 21,291 leaf-node classes across 8 visual domains, with 32,541 evaluation instances whose ground-truth labels cover 15,479 categories across four complementary tasks: closed-set recognition, contrastive selection, precise localization, and open-world naming, covering different directions of knowledge retrieval and varying degrees of answer openness to comprehensively evaluate visual world knowledge. Most questions are constructed from taxonomically proximate categories, forcing precise discrimination among fine-grained entities. Experiments reveal that the strongest closed-source model achieves only 80.4% average accuracy, exposing capability gaps hidden under conventional benchmarks. We further construct VisualWorldBench-50K, a training resource covering 50K+ categories. Fine-tuning on VisualWorldBench-50K yields over 21 percentage points improvement on VisualWorldBench, with consistent transfer to public fine-grained recognition and visual knowledge benchmarks.
Read the full abstract

Discrete Mathematics

Routes planning in a condominium company using mathematical optimization

Lara Rocha and Pedro Martins

This work focuses on planning daily routes for a group of employees of a condominium company. Given a set of employees and a set of customers with pre-scheduled visits, we want to build doubly-open routes to guarantee visits to all customers, ensuring that each customer is visited by an employee with the appropriate skills to carry out the required activity. The travel times between customers and the service time in each customer are known. The problem includes time windows for visiting customers, each employee has a maximum working time, and the daily working period is between 9 AM and 6:30 PM. The routes respect the lunch break period if they extend from morning to afternoon, which means that lunch break can be avoided if the route is fully completed in the morning or in the afternoon. The aim is to minimize the distance covered by all routes with a penalty for including lunch breaks. The distances and times between customers are collected via an API to Microsoft Azure; and the solutions are then displayed on Bing Maps. The results obtained show the operational advantages of an efficient and effective vehicle routing planning in a specific condominium management company. The model can be extended to other service provider companies.
Read the full abstract

Data Structures and Algorithms

Undirected Network Coding and Faster External-Memory Integer Sorting

Baochun Li, Zongpeng Li and Sirui Liu

Network coding allows intermediate nodes to combine messages, whereas routing only forwards them. The Li-Li conjecture asserts that network coding and fractional routing have the same capacity region in undirected multiple-unicast networks. We give a counterexample to this conjecture and apply it to external-memory integer sorting. Since Aggarwal and Vitter posed the question in 1988, whether integer sorting can beat the external-memory comparison bound has remained a central and extensively studied open problem. We give a deterministic sorting algorithm: for $N$ integers of $\Theta(\log N)$ bits, a memory holding $M$ integers, and block transfers of $B$ integers, where $2B\le M\le N$, the number of block transfers is $O\left(\frac{N}{B}\left[1+\frac{\log N}{(\log(M/B))^{1-\varepsilon}(\log M)^\varepsilon}\right]\right)$, where $\varepsilon>0$ is an absolute constant. In a range of parameter regimes, this bound is asymptotically smaller than the classical comparison-sorting bound. We also use Lean 4 to verify the sorting algorithm's correctness and complexity bounds, and the strict advantage of coding over routing.
Read the full abstract

Machine Learning

A Large-Dimensional Analysis of Least-Squares Federated Learning via Generalized Total Variation Minimization

Ekkehard Schnoor, Malik Tiomoko and Alexander Jung

Federated learning over decentralized networks inherently trades off the suppression of local finite-sample noise through collaboration against the bias that collaboration introduces between heterogeneous nodes. We study this tension by performing a large-dimensional analysis of Generalized Total Variation Minimization with least-squares models. Using tools from random matrix theory, we show that Lipschitz observations of the global GTVMin solution concentrate in large dimensions, universally for any underlying network structure. We provide exact analytic expressions for the asymptotically precise performance under both i.i.d. and non-i.i.d. Gaussian mixture distributions on a fixed network. For the latter, these expressions quantify how regularization balances local adaptation against collaboration. Our theoretical results are confirmed by experiments on synthetic and real-world data.
Read the full abstract

Computer Vision

Feasibility of a Patient-Specific Digital Twin for Ventriculoperitoneal Shunt Trajectory Planning

Kritsanavis Chongsrid, Vich Yindeedej, Supaporn Konmun et al. (+2 more)

Catheter obstruction is the commonest cause of shunt failure; freehand insertion at Keen's point leaves the tip free in cerebrospinal fluid in only one case in ten. We built a digital twin that reconstructs skull, scalp, ear and ventricles from preoperative CT by deterministic geometry, searches every entry around Keen's point for the perpendicular trajectory with the longest intraventricular length under safety constraints, and returns the plan as two ruler measurements from the ear and two depths. Executed freehand in 11 consecutive adults (IDEAL Stage 2a), it achieved Hayhurst Grade 1 placement in 10 of 11 against 22 of 213 historically (p = 9 x 10^-9), median entry and angular deviations 2.5 mm and 5.9 degrees, 1.4 cm more intraventricular catheter than Keen's point, and no intraoperative complication. The technique was unchanged; the gain came from choosing entry and depth per patient. Whether it lowers revision is the next stage's hypothesis.
Read the full abstract

All authors Kritsanavis Chongsrid, Vich Yindeedej, Supaporn Konmun, Xulei Yang, Si Yong Yeo

Computer Vision

GDBA-Net: A Gated Dual-Branch Adaptation of SAM2 for Robust Cross-Domain Nuclei Segmentation

Biplab Poudel, Betserai Albert Magarire, Anish Kumar Shah et al. (+2 more)

Accurate nuclei segmentation underpins quantitative computational pathology, yet remains challenging due to morphological variability, weak boundaries, nuclear crowding, and pronounced domain shifts arising from differences in tissue preparation, staining, and acquisition. While deep learning and recent foundation models have improved segmentation performance under matched distributions with the original training data, their generalization across heterogeneous histopathology domains remains limited. We introduce GDBA-Net, a parameter-efficient nuclei segmentation framework built around the proposed Gated Dual-Branch Adapter (GDBA). GDBA-Net adapts a frozen SAM2 encoder by integrating pretrained semantic representations with explicit frequency-derived structural priors. The complementary representations are independently transformed through Grid MLP branches and dynamically integrated using sample-adaptive channel-wise gating, selectively emphasizing structural cues relevant to nuclear delineation. Despite using only 4.3 million trainable parameters, GDBA-Net achieves state-of-the-art in-dataset performance and consistently superior cross-dataset generalization across six diverse nuclei segmentation benchmarks under fully supervised, cross-dataset, and five-shot adaptation protocols, outperforming competing CNN-, transformer-, and foundation model based methods. Comprehensive ablations confirm that these gains arise from the coordinated design of semantic-frequency complementarity and adaptive gating rather than increased model capacity. By uniting transferable semantic representations with explicit structural priors, GDBA-Net provides a robust, annotation-efficient solution to domain variability in nuclei segmentation. https://github.com/biplabpoudel25/GDBA_Net
Read the full abstract

All authors Biplab Poudel, Betserai Albert Magarire, Anish Kumar Shah, Jiacheng Xie, Dong Xu

Machine Learning

Rethinking Prediction Horizons: Decoupling Model Queries from State Updates in Few-Step MeanFlow

Tinglan Wang, Haixu Liu, Shaobo Chen et al. (+4 more)

MeanFlow predicts interval-average velocities, but conventional few-step sampling ties the interval queried from the model to the duration of the state update. We separate these two quantities at inference, fixing the model, update grid, and number of queries while varying the prediction horizon. Our central finding is that the horizon response depends on both query position and the states entering that query. Across seven independently trained CIFAR-10 checkpoints at classifier-free guidance (CFG) scale 2, a longer horizon worsens FID at early steps but improves it at the midpoint. To probe whether query time alone determines this response, we fix the midpoint query and current/subsequent guidance while changing only earlier CFG. In held-out factorial outcomes on five previously studied checkpoints, the mean midpoint horizon benefit increases from 0.109 to 1.308 FID, indicating sensitivity to the incoming-state distribution. A frozen four-query rule reduces mean CIFAR-10 FID by 2.023 without retraining or extra queries, including gains on five checkpoints trained after rule selection. A milder rule also improves mean FID by 1.041 on a stronger frozen grid, with fixed-rule FID gains on the evaluated public ImageNet-256 and AFHQ checkpoints. However, matched-budget calibration favors query-start shifting in historical CIFAR tests, and the tested Horizon allocation adds no FID benefit beyond Grid+Start on two public ImageNet models. Thus, prediction horizon is a distinct, position- and state-sensitive inference control, but not a universally preferred one.
Read the full abstract

All authors Tinglan Wang, Haixu Liu, Shaobo Chen, Nuo Qian, Qi Gao, Kai Wen, Yu Yao

Networking

SPARC: Substation-to-Peering Assessment of Resilience and Cascades

Tanya Shreedhar, Kevin Vermeulen, Amreesh Phokeer et al. (+2 more)

Internet Exchange Points (IXPs) operate from colocation facilities that draw power from regional grids and store limited backup fuel. In April 2025, the Iberian blackout kept the Spanish grid down for 16 hours, beyond the 12 hours of fuel required by the industry standard, while traffic at major Iberian IXPs fell by up to 70%. Yet operators cannot tell which facilities a grid event cuts off, whether it outlasts their backup, or where traffic from a failed IXP goes. We present SPARC, which maps power-grid and Internet dependencies from substations to IXPs and their traffic. Using open data validated against transmission operators, SPARC links 552 facilities hosting 275 IXPs to the substations that likely feed them. We spread grid events over this map and measure failover through controlled BGP withdrawals on the PEERING testbed. Disabling the most-shared 16% of facility-feeding substations cuts power to half of all IXP presences (an IXP at one facility), under half of what randomized maps require. Losing the 5% most central substations cuts power to 0.1% of IXP presences directly but to 9-52% once the grid splits into disconnected islands. Traffic often fails over to nearby sites that share the failing grid: a cascade affecting four metropolitan areas sends up to 40% of one area's monitored addresses to another of the four. We introduce SPARC-Shield to select substations to harden or facilities whose backup to extend. Against an attacker disabling ten substations, hardening the ten selected by SPARC-Shield reduces offline IXP presence by 72%, compared with 52% for the best standard ranking.
Read the full abstract

All authors Tanya Shreedhar, Kevin Vermeulen, Amreesh Phokeer, Italo Cunha, Fernando Kuipers

Machine Learning

Trainable at Depth, Solvable at Small Size: What the Loss Surface of a Polynomial Activation Dictates for Designing, Initializing and Training Networks

Xiuyi Yang

What makes a deep network train, and what shape is its loss surface? We answer both with one construction at two scales: a polynomial activation f (z) = z + {\lambda}z 3 , clamped and projected onto the unit sphere, plus an unnormalized identity path. As a deep stack it trains to one hundred layers on MNIST with no normalization machinery; as a small model its square loss is a polynomial system we solve exactly. The small scale explains the large one: it prescribes how to design, initialize and train the stack.
Read the full abstract

Software Engineering

The Code of Theseus: Can Agents Rebuild Software from Recovered Design Documents?

Cezar Sas, Sandra Mitrović, Fabio Barth et al. (+2 more)

LLM agents increasingly design a system before implementing it, yet the data to train and evaluate them centres on implementation and testing: requirements and design artefacts are often missing, outdated or private. Source code, in contrast, is abundant. Prior work recovers architecture documentation by feeding LLMs the source code or letting agents explore it, which saturates the context, does not scale to many repositories, and yields a single artefact rather than aligned ones. We present a framework that recovers aligned stakeholder intent, requirements and architecture from a compact repository representation, a static dependency graph and key contextual files, and release the artefacts of 10,764 Java, Python, C/C++ and C# repositories as a dataset. An LLM judge rates the artefacts of three generators as clear and internally consistent, but less reliably correct and at the right level of abstraction. To test whether the artefacts carry the design, we pose a software Ship of Theseus: a coding agent rebuilds 29 Commit0 libraries from the intent alone or with the recovered requirements and architecture, given only an interface contract and part of the tests. The artefacts do not significantly change the held-out pass rate (33.7% vs. 35.8%), but bring the rebuilt module structure significantly closer to the original (module F1 from 0.54 to 0.63), with either generator and even with concise artefacts. Recovered artefacts thus convey design that tests do not capture, offering a scalable source of design-level data for agents that design before they implement, and the round trip measures this without human references.
Read the full abstract

All authors Cezar Sas, Sandra Mitrović, Fabio Barth, Georg Rehm, Matteo Salani

Computation and Language

Which Knowledge Should LLMs Trust? Uncovering a Generalizable Steering Direction Across Knowledge Conflicts

Yibin Lou, Yi Hu, Jianhui Chen et al. (+3 more)

Large Language Models (LLMs) frequently encounter conflicts between internal parametric knowledge and external contextual knowledge, yet the mechanisms governing this selection remain unclear. We show that the choice between context and parametric knowledge is linearly represented and can be controlled along a single direction in the model's residual stream. We first demonstrate that this direction generalizes across factual-conflict settings, including context formats, knowledge categories, and datasets, while consistently shifting behavior between parametric and contextual answers. We then transfer a direction learned from simple factual conflicts to three more complex task families: RAG, memory, and rule following, where models must acquire information from retrieved documents, interaction histories, or explicit rules. Across four model families and multiple transfer settings, steering consistently improves the use of prompt-provided information over the unsteered model. These results suggest that the direction captures a transferable preference for external knowledge rather than a mechanism specific to factual conflicts.
Read the full abstract

All authors Yibin Lou, Yi Hu, Jianhui Chen, Chenpeng Wang, Muhan Zhang, Liangming Pan

Game Theory

No-Regret Learning in Stackelberg Security Games with an Unknown, Bounded Set of Attacker Types

Sabil Ahmad, Ishita, Aditi Singh et al. (+1 more)

In a repeated Stackelberg security game, a defender commits in every round to a randomized allocation of its resources, and an attacker, drawn from a set of types, best-responds. Existing no-regret guarantees assume that the defender knows this set of types in advance. We drop this assumption: the set of types is unknown, and its size is at most $K_{\max}$, an upper bound that the defender is not told. We show that what can be learned then depends on what the defender observes. If it observes the type of each attacker, an algorithm that adds types to its list as they appear achieves regret $\tilde O(nK_{\max}\sqrt{T})$ over $T$ rounds with $n$ targets, the same $\sqrt T$ rate as with known types, at an extra factor of only $\sqrt{K_{\max}}$ and without ever using $K_{\max}$. If it observes only the attacked target and types come from an infinite universe, no algorithm has sublinear regret: in a fixed three-target game every algorithm suffers regret at least $T-2$, even against a single attacker type, because it cannot tell which type it faces. If, in addition, it is told every $B$ rounds which types have appeared so far, sublinear regret returns, at $\tilde O(nK_{\max}^{4/3}T^{2/3}+K_{\max} B)$. The first term is the rate of the partial-information learner we use, and we show that the additive cost of the delay cannot be avoided.
Read the full abstract

All authors Sabil Ahmad, Ishita, Aditi Singh, Swaprava Nath

Robotics

FoveaVLA: Foveation Actions Replace the Wrist Camera in VLAs

Bingao Chen, Lihan Zha, C. Karen Liu et al. (+1 more)

We present FoveaVLA, a Vision-Language-Action (VLA) model that replaces the wrist camera with learned foveation actions. Wrist cameras provide close-up detail but restrict the view to the end-effector's surroundings. External cameras see the whole scene, but the policy receives them at low resolution, so the fine detail needed for manipulation is lost. FoveaVLA recovers this detail by predicting, alongside each action chunk, the interaction region to observe at the next policy query. The policy then receives crops of these regions, taken from the original-resolution images, together with the full views. We build FoveaVLA on $\pi_{0.5}$ in two configurations, Single and Dual, using one or two external stereo cameras with full and foveal RGB and depth inputs, and post-train both on DROID with RoboInter interaction annotations. Across four real-world task suites, FoveaVLA-Single achieves 70.8% success with one external RGB-D view, and FoveaVLA-Dual achieves 77.1%, versus 52.1% for the wrist-equipped $\pi_{0.5}$. Both match the baseline on plain pick-and-place and improve success with distractors, cloth manipulation, and far-away objects.
Read the full abstract

All authors Bingao Chen, Lihan Zha, C. Karen Liu, Guy Tevet

Data Structures and Algorithms

Towards Dynamic Top-$k$ Simple Shortest Paths

Gianlorenzo D'Angelo, Andrea D'Ascenzo, Mattia D'Emidio et al. (+2 more)

Computing the top-$k$ simple shortest paths between two vertices of a graph, i.e., solving the P-KSSP problem, is a fundamental primitive in graph analytics. In modern scenarios, however, queries are rarely one-shot: parameters change as applications adapt their information needs, and recomputing a solution from scratch at every request quickly becomes a bottleneck. We study P-KSSP in a dynamic setting, focusing on changes of the parameter $k$: given a top-$k$ solution for a pair of vertices and a new value $k'$, the goal is to compute a top-$k'$ solution reusing the previously computed paths. Our approach rests on the pseudo-tree, a compact representation of a collection of simple $s$–$t$ paths: we show that building it from an arbitrary, unstructured solution amounts to trie construction, and give two constructions running in time linear in the size of the collection. On top of it, we design UPD, a dynamic algorithm that updates a solution upon a change of $k$; we prove its correctness, analyze its complexity, and show that it is a general, solver-agnostic scheme, which starts from the paths alone, is parametric in the routine that generates candidate paths and combinable with known preprocessing techniques, as well as with an online pruning rule that we formalize. While a worst-case update is asymptotically comparable to recomputing from scratch, an extensive experimental evaluation shows that a variant of UPD is the fastest method in the large majority of tested configurations, often by a wide margin (with speedups over recomputation that grow with $k$ and $k'$ and exceed $60\times$), yielding a practical recipe for the configuration to deploy.
Read the full abstract

All authors Gianlorenzo D'Angelo, Andrea D'Ascenzo, Mattia D'Emidio, Gabriele Di Stefano, Marta Innamorati

Artificial Intelligence

Beyond Correctness: Compute-Responsive Data Selection for Reasoning Model Post-Training

Liang He, Xiangyu Li, Yuxin Wang et al. (+4 more)

Reasoning language models increasingly rely on test-time compute scaling, yet most existing work studies how to allocate inference computation while treating the model itself as fixed. We ask a complementary question: can reasoning models be post-trained to make better use of a given amount of test-time compute, and which examples are most useful for this goal? Our key observation is that reasoning problems differ substantially in their response to additional computation. Some are already solved under limited compute, others remain unresolved even after further reasoning, while a distinct subset changes from complete initial failure to success when additional reasoning is provided. We call these compute-responsive (CR) examples and introduce Compute-Responsive Data Selection (CRDS), a post-training data-selection principle that prioritizes examples with empirically demonstrated marginal utility from additional reasoning. On Qwen3-8B, SFT on CR examples recovers 45.3% of held-out low-compute failures, compared with 40.6% when training on already-solved examples. The advantage extends beyond the selected CR split: SFT-CR reaches 51.8% overall accuracy on an independently sampled test distribution, with the gain concentrated on compute-responsive problems. CR training also shifts the compute–accuracy frontier across inference budgets from 512 to 4,096 tokens, outperforming matched non-CR training throughout and reaching 48.1% recovery at the largest budget. The same pattern is supported across multiple data sources, remains visible under on-policy distillation, and extends to Llama-3.1-8B-Instruct when CR sets are reconstructed separately for that model. Better test-time compute utilization can therefore be learned through post-training, with recoverable responses to additional reasoning providing a particularly informative signal for selecting training examples.
Read the full abstract

All authors Liang He, Xiangyu Li, Yuxin Wang, Hanbin Zhang, Mingyu Cao, Zhaomiao Guo, Xilu Wang

Computer Vision

GeoCloak: An Agentic Framework for Semantics-Preserving Geolocation Privacy via Peripheral Clue Manipulation

Yixin Wu, Mingjie Li, Yun Shen et al. (+2 more)

The rapid advancement of large vision language models (LVLMs) has made geolocation inference from images highly accurate and accessible. This poses serious privacy risks: a passive adversary with black-box LVLM access can infer a user's location from a single photo with no metadata. Existing protection methods rely almost exclusively on adversarial perturbation, which introduces perceptible visual artifacts and utility degradation. We identify the visual-geolocation asymmetry: geolocation-decisive clues, such as street signs, tend to be visually peripheral but dominate location inference. Exploiting this asymmetry, we propose GeoCloak, an agentic framework for semantics-preserving geolocation privacy protection. It acts as a clue-aware planner that analyzes each image to identify geolocation-relevant clues, dispatches targeted semantic edits to directly suppress peripheral clues, and autonomously configures adversarial perturbation to further suppress residual geolocation features, all under iterative utility-aware feedback. We evaluate GeoCloak on four diverse datasets spanning 108 countries and 576 cities against five LVLMs and three inference methods. GeoCloak reduces overall city-level accuracy from 41.2% to 9.9%, approximately half of the best-performing baseline, while achieving substantially superior utility preservation both qualitatively and quantitatively. We hope this work takes a meaningful step toward practical, semantics-preserving geolocation privacy protection, giving users greater control over the location information inadvertently embedded in their shared images.
Read the full abstract

All authors Yixin Wu, Mingjie Li, Yun Shen, Michael Backes, Yang Zhang

Machine Learning

GBTSVR: Twin Support Vector Regression via Granular Ball Computing

A. Quadir and M. Tanveer

Regression models used in real-world applications are often challenged by noise, outliers, data heterogeneity, and asymmetric error characteristics, which severely limit the effectiveness of conventional point-wise learning models. While twin support vector regression (TSVR) partially addresses these issues through the use of nonparallel regression bounds, it still relies on individual samples as basic learning units. As a result, TSVR remains highly sensitive to noisy and contaminated observations and fails to leverage the underlying structural and regional information inherent in complex datasets, leading to suboptimal robustness and generalization performance. To address these limitations, we introduce granular ball twin support vector regression (GBTSVR) and its quantile extension, granular ball twin support vector quantile regression (GB-TSVQR). The proposed models integrate granular ball computing with TSVR, replacing individual data points with compact, structure-preserving granular balls as fundamental learning units. This region-level representation effectively suppresses noise and outliers while preserving intrinsic data geometry, thereby enhancing robustness, stability, and generalization. GBTSVR constructs nonparallel lower- and upper-bound regression functions over granular balls, significantly reducing sensitivity to sample-level perturbations. Building upon this framework, GB-TSVQR further incorporates quantile regression principles to simultaneously capture heterogeneous trends and asymmetric error distributions across different quantile levels. Extensive experiments on synthetic datasets, time-series benchmarks, and outlier-contaminated data demonstrate that the proposed GBTSVR and GB-TSVQR consistently outperform state-of-the-art regression models.
Read the full abstract

Computation and Language

What Is a Clarification? Estimating Model-Dependent Input Ambiguity

Jingqin Tang, Tim Tomov, Sebastian Schmidt et al. (+2 more)

Ambiguous inputs can admit multiple interpretations, each potentially yielding a different answer. Existing work largely treats this ambiguity as a property of the input, yet for a particular language model, only some alternatives may be known and answerable. We therefore distinguish world ambiguity, capturing all alternatives admitted by an input, from model ambiguity, capturing those accessible to the responding model. To characterize these alternatives, we formalize natural-language clarifications through three desiderata: uniqueness, faithfulness, and answer equivalence. We show that these desiderata induce a one-to-one correspondence between valid answers and equivalence classes of clarifications, implying that any interpretation can be represented either by its answer or by the clarifications that uniquely express it. We leverage this correspondence to introduce \circleclar{}, a black-box method that estimates the set of answers available to a model by checking consistency along an answer–clarification–answer cycle. Across open-domain QA and executable Text-to-SQL, \circleclar{} achieves strong recovery of these answer sets. In QA, the recovered support yields the strongest overall performance in model-dependent ambiguity detection and enables effective selection among answering, clarifying, and abstaining. Together, these results demonstrate the practical value of a model-dependent view of ambiguity and its characterization through an answer-clarification correspondence.
Read the full abstract

All authors Jingqin Tang, Tim Tomov, Sebastian Schmidt, Dominik Fuchsgruber, Stephan Günnemann

Computation and Language

FACES: Fine-grained Gated Semantic Alignment for Multimodal Emotional Support Conversation

Hang Ren, Peidong Wang, Yichen Gao et al. (+4 more)

In recent years, with advances in multimodal interaction technologies and the growing demand for intelligent emotional support, Multimodal Emotional Support Conversation (MESC) has emerged as an important task. However, existing systems still suffer from shallow responses and inappropriate support strategies due to weak semantic feedback, cross-dimension reward compensation, and coarse reasoning supervision. To address these challenges, we propose FACES, a fine-grained gated semantic alignment framework for MESC. Specifically, FACES integrates multi-source rewards to construct multidimensional semantic and functional alignment signals, employs a non-compensatory gated fusion mechanism to reduce cross-dimension reward offset, and leverages structured Chain-of-Thought (CoT) anchors to perform token-level credit assignment. Under a controlled response-length protocol, FACES improves the average MLLM-based evaluation score from 3.61 to 4.04 and achieves an 80.22% pairwise win rate over SOTA. Human evaluation under the same protocol yields a 75.3% preference rate. Results on three external datasets further support transferability. The code and dataset are available at https://github.com/OCT-ARR/arr.
Read the full abstract

All authors Hang Ren, Peidong Wang, Yichen Gao, Wen Zhang, Jiayang Yu, Xiaocui Yang, Shi Feng

Cryptography and Security

CoRedSkill: Stealthy Skill Composition Attacks via AST-Guided Injection

Qichen Zhao, Shengfang Zhai, Qingni Shen et al. (+2 more)

Agents extend their capabilities through reusable skills that package instructions, executable code, and resources. Composing these skills enables complex task workflows, but their dependencies can also connect malicious operations across skills, making their combined intent difficult to identify through individual-package auditing. We introduce CoRedSkill, a compositional red-teaming method that forms stealthy attacks through cooperating skills while preserving their original utility. CoRedSkill decomposes an existing workflow into interdependent skills and embeds attack operations in task functions through abstract syntax tree (AST) rewriting, with skill instructions guiding the agent to autonomously select and invoke the relevant skills, and workflow dependencies and intermediate artifacts connecting the injected operations during task execution. We construct a benchmark of 150 cases covering 50 tasks and three attack objectives to evaluate attack success, task completion, and audit detection. In the main evaluation, CoRedSkill achieves 98.0% attack success and 95.3% task success with a 3.3% detection rate, and a joint success rate of 90.7% (attack success, task completion, and audit evasion in the same case) compared with 6.7–42.7% for the four baselines. Skill compositions can therefore complete legitimate tasks while carrying out unauthorized operations that evade auditing.
Read the full abstract

All authors Qichen Zhao, Shengfang Zhai, Qingni Shen, Zhonghai Wu, Jiaheng Zhang

Information Retrieval

Reading the Neighborhood: Text-Attentive Graph Reranking for Automated Subject Indexing

Lakshmi Rajendram Bashyam and Ralf Krestel

Subject indexing assigns descriptors from a controlled vocabulary to bibliographic records, a task characterized by thousands of labels, few positive examples per label, and a long tail of rarely used descriptors. Existing extreme multi-label classifiers treat the vocabulary as a flat list, ignoring the semantic relations that thesauri encode. We present TAGHN, a heterogeneous graph reranker over a per-document subgraph of retrieved candidates augmented with the thesaurus neighborhood. The relevance of a candidate descriptor is estimated jointly from its similarity with the document and also the degree to which the candidate's thesaurus neighborhood matches the document content. A decomposed scoring head combines interpretable graph-derived features with retrieval signals. On STW, an economic thesaurus of 6,198 descriptors, TAGHN reaches 0.380 macro F1 against 0.346 for a fine-tuned cross-encoder and 0.278 for a fixed-vocabulary XMLC classifier. In terms of micro F1, it outperforms reranking models and stays competitive (-3.0) with the best fixed-vocabulary XMLC model. Qualitative analysis shows the graph resolving errors that per-label scoring cannot, selecting a narrower gold descriptor where an unstructured reranker settles for its broader parent.
Read the full abstract

Cryptography and Security

ESAU: Expeditious Swarm Remote Attestation for Pub-Sub IoT Networks with Untrusted Cloud-based Proxy Verifier

Ruei-Hau Hsu, Yu-Hsiang Hu, Rafael Kaliski et al. (+1 more)

As Internet of Things (IoT) devices collect and process data, security is essential. IoT devices often combine with other emerging technologies, such as artificial intelligence (AI) and fifth-generation (5G) communications, to support applications such as smart grids and internet-of-vehicles (IoV). Because IoT devices are deployed remotely, remote attestation (RA) effectively verifies each device's state. However, the scale of IoT deployments makes device management challenging for owners. Therefore, cloud-assisted IoT management services often provide RA solutions. Although cloud-based IoT services can securely manage IoT devices, passive malicious administrators or hacked IoT device owners' accounts can control IoT devices without cloud IoT service authorization. Therefore, this work proposes a cloud-assisted RA system that addresses three main challenges of IoT device RA: 1) lightweight RA design for resource-limited IoT devices, 2) unauthorized access to IoT devices by passive malicious administrators or compromised owner accounts on the cloud IoT management platform, and 3) RA may not extend to an authenticated key exchange protocol. In addition to proposing a swarm RA scheme that addresses these three challenges, this work analyzes the system's security properties and evaluates its feasibility through implementation and simulation.
Read the full abstract

All authors Ruei-Hau Hsu, Yu-Hsiang Hu, Rafael Kaliski, Tse-Yuan Yu

Machine Learning

Sequential Learning-Based Physics-Informed Neural Network Modeling for Parameter Estimation on a Power Converter

Samuela Rokocakau, Panagiotis Kakosimos, Giulia Tresca et al. (+1 more)

A staged physics-informed neural network (PINN) is proposed for identifying the inductance, capacitance and parasitic resistances of a DC buck converter from time-domain data. Gradients are released in the order of L to C to rL to rC, a sequence derived from sensitivity analysis. The five-stage schedule recovers L and C to within 5% and the series ESR to 10%, while a three-stage variant attains the same accuracy 11% faster. An 84 factorial sweep confirms that the trained PINN remains accurate for simultaneous 8% perturbations of all four components, achieving a waveform RMSE of 2.6 x 10-2 over 5 ms-an order-of-magnitude better than a simultaneous-training baseline.
Read the full abstract

All authors Samuela Rokocakau, Panagiotis Kakosimos, Giulia Tresca, Pericle Zanchetta

Software Engineering

SWE-Nexus: Automated Composition of Software Engineering Tasks from Test Execution

Songwei Yu, Ziqi Miao, Jingxuan Xu et al. (+10 more)

Historical code changes provide convenient task boundaries, but do not identify which production functions selected tests exercise or support composing validated behaviors beyond a single change. We present SWE-Nexus, a fully automated pipeline that constructs repository-level implementation tasks from test execution. Four construction agents select tests, adapt tracing and evaluation protocols to each repository, and generate behavioral specifications. For each selected test file, tracing records an observed execution footprint of production function bodies. A scope is retained only if all expected scoring points pass in the original repository and masking the traced bodies causes at least one failure or error. Compatible scopes are composed deterministically, with shared functions counted once and every constituent test file revalidated under the complete combined mask. The benchmark contains 600 tasks from 105 repositories across six programming languages, including 116 atomic and 484 composite tasks. We evaluate 12 models under three agent scaffolds. The best configuration resolves 37.8% of tasks in one attempt, and scope-weighted scoring measures partial recovery. SWE-Nexus enables task construction from validated behavior beyond individual historical changes.
Read the full abstract

All authors Songwei Yu, Ziqi Miao, Jingxuan Xu, Junhao Fan, Haoyang Huang, Yuan Xiong, Zihang Xu, Weihao Li, Wenqiang Zhu, Yuechen Luo, Xinshun Feng, Yanan Wu, Ken Deng

Systems and Control

Topology-adaptive Estimation of Maximum Voltage Unbalance for Distribution Networks with Finite-sample Guarantee

Jinbao Wang, Jun Liu, Shiliang Zhang et al. (+2 more)

High penetration of single-phase distributed power production and frequent topology changes can significantly increase voltage unbalance in distribution networks. Reliable probabilistic power flow (PPF) assessment for the network, therefore, must account for topology-dependent power-flow equations, numerical error in the voltage unbalance factor (VUF), and finite-sample uncertainty in power measurements. In this paper, we propose Topology-Adaptive Certified Quantile Estimation (TACQE) to estimate the maximum VUF across monitored electric buses. Following a local network change, TACQE updates the cross-couplings among power-flow sequences, and certifies candidate PPF solutions against the updated nonlinear equations. The resulted power-flow state enclosures provide rigorous bounds on the maximum VUF. Furthermore, we derive independent reference-solver evaluations to calibrate discrepancies from the reference PPF solution. This leads to a uniform sampling bound that yields simultaneous upper-tail probabilities and quantiles, which provides a finite-sample guarantee for the estimated maximum-VUF distribution over an entire range of VUF thresholds. From our numerical simulations across 101 fixed-topology quantiles, it demonstrates that the designed TACQE can provide a fast candidate post-change power flow solution where it reuses most of the previous network model and modifies only the affected bus blocks.
Read the full abstract

All authors Jinbao Wang, Jun Liu, Shiliang Zhang, Xuehui Ma, He Zhang

Machine Learning · Cross-listed from stat.ML

Weighted Adjoint Matching: Finetuning Flow Matching Models with Time-Dependent Control Cost

Eliot Beyler and Francis Bach

We introduce Weighted Adjoint Matching, a method to finetune a pretrained diffusion or flow matching model given a differentiable reward. We derive it from Adjoint Matching, which reformulates finetuning as a stochastic optimal control problem with a quadratic control cost. Building on the observation that these models generate coarse features before fine details, we place a time-dependent weight on the control cost. A high cost early in the generative process preserves balance and diversity in the coarse features, which carry most of the semantics, while a low cost at the end lets the model increase the reward by refining details. We prove that our method is equivalent to maximizing the reward under time-varying Kullback–Leibler regularization on the generative paths. Experiments on Gaussian mixtures and on MNIST show a better tradeoff between reward and mode balance than Adjoint Matching.
Read the full abstract

Distributed Computing

TrimSO(2): Fast and Memory-Efficient Training of Equivariant Interatomic Potentials

Chen Wang, Yuanchang Zhou, Guangming Tan et al. (+2 more)

Equivariant interatomic potentials enable accurate and efficient simulations of molecules and materials. Among them, models based on SO(2) convolutions reduce the cost of tensor products and achieve state-of-the-art accuracy, but their training remains time-consuming and memory-intensive. Existing training implementations materialize large edge tensors for representation transformations surrounding dense matrix multiplications, increasing memory usage and data movement. We observe that these transformations can be integrated into the model's existing dataflow, avoiding separate intermediate materialization. We introduce TrimSO(2), a fast and memory-efficient algorithm for SO(2)-based equivariant models. It realizes this principle through rotation on gather, recombination with activation, and inverse rotation on reduction, with support for the higher-order differentiation required for training. By avoiding intermediate materialization, TrimSO(2) delivers 2.71-16.73$\times$ operator speedups across forward, backward, and double backward and reduces model-level high-bandwidth memory traffic by 3.81-9.88$\times$. Finally, TrimSO(2) achieves 2.07-2.94$\times$ training speedups and 1.53-3.16$\times$ reductions in peak training memory over PyTorch implementations for official eSEN, UMA, and EquiformerV3 workloads on A100 and H100 GPUs.
Read the full abstract

All authors Chen Wang, Yuanchang Zhou, Guangming Tan, Mingzhen Li, Weile Jia

Software Engineering

LintEval: Benchmarking LLMs on Linter-Guided File Refactoring

Tongwei Deng, Jinyu Zhu, Yue Cao et al. (+3 more)

Linters are widely used in code maintenance and real-time IDE editing to surface lint alerts related to performance, style, error-prone, and other patterns in source files. However, existing code benchmarks do not explicitly evaluate linter-guided file refactoring, where models must reduce lint alerts in source files with correct initial functionality while satisfying functionality preservation. We introduce LintEval, a multilingual benchmark for linter-guided file refactoring across C++, Python, and Java, comprising 13,061 programs with correct initial functionality and 171,662 lint alerts extracted using language-specific linters. LintEval jointly evaluates functionality preservation, lint alert reduction, new alert avoidance, and edit minimality. Experiments reveal a persistent trade-off: closed-source commercial models reach up to 90.4% Reduction@P but achieve only 67.1%-68.3% Pass@1, while several open-source small models show low or negative Reduction@P. A controlled fine-tuning study further shows that supervision format matters: refactoring-oriented SFT raises Qwen2.5-Coder-7B Reduction@P from 24.7% to 76.0%. LintEval enables controlled evaluation of models that improve source files under functionality preservation.
Read the full abstract

All authors Tongwei Deng, Jinyu Zhu, Yue Cao, Chenxu Wang, Yanzhen Zou, Bing Xie