Robotics
Switched Turn-based Adaptive Source Seeking Strategy using Estimation and Information-driven Direction of Improvement
Shubhra Banerjee and Satadal Ghosh
Source seeking arises in applications such as gas leak localization, radiation monitoring, and environmental surveillance, where the origin of an unknown signal field must be estimated from spatial measurements. In practice, the source location is not directly observable and must be inferred from noisy scalar measurements collected during motion.In robotic source seeking, estimation and motion are closely linked: measurements improve the source estimate, while the chosen trajectory affects the quality of future measurements.Existing loop-based
geometric strategies generate feasible motion but do not explicitly use estimation uncertainty to regulate direction updates.This paper presents a loop-based source-seeking framework that combines Extended Kalman Filter (EKF) estimation with Fisher Information
Matrix (FIM)-based direction selection. The source estimate is updated during motion, and the heading is changed at loop boundaries using both estimation uncertainty and predicted information gain. A measurement-based stopping condition is used to detect convergence without requiring prior knowledge of the source location.Simulation results under stationary and moving source scenarios demonstrate improved tracking performance and reduced estimation error compared to purely information-driven or estimate-driven strategies.
Read the full abstract ▾
Computation and Language
Signal or Noise? A Benchmark Study of Agent Skills in Web Development
Ziyue Yang and Fan Ding
Agent Skills are reusable procedural modules that are increasingly injected into coding-
agent sessions to encode framework conventions, anti-patterns, and reusable tools. However, because each injected Skill expands the
prompt of every query, an effective Skill
benchmark must determine not only whether an
agent can solve a task, but whether the Skill should have been injected at all. We introduce WebDev-Skills-Bench and use it for a controlled empirical study of 31 public WebDev Skills on 50 Web-Bench projects and 1,000 ordered tasks. The
benchmark compares four matched conditions, including a length-matched irrelevant control and leave-one-out component ablations. To isolate Skill effects from
prompt-length artifacts, we place only SKILL.md in the
prompt while mounting auxiliary files into the
agent workspace. Across four models, target Skill injection reduces mean Pass@2 by 1.3% to 4.2%, lowers task completion depth, and increases
token cost by 72% to 394%, with gains in only 17% to 36% of Skill-project pairs. Length-matched controls reveal two failure modes: some models are length-distracted, where an equally long irrelevant Skill reproduces most of the loss, while others are content-misled, where
prompt length is neutral but Skill content still lowers Pass@2 by 1.1% to 1.4%. Further analysis shows that losses concentrate on easy early tasks, Skill rankings transfer weakly across models, and anti-pattern rules outperform example-heavy content within helpful Skills. These findings recast a matched Skill as a hypothesis about a particular Skill-project-model triple rather than a portable asset, reframing injection as a per-deployment routing decision and making length-matched controls and per-model audits a minimum standard for
Agent-Skill evaluation.
Read the full abstract ▾
Computer Vision
Cultural Moment Benchmark: Evaluating Video Cultural Reasoning and Grounding in Southeast Asia
Burak Satar, Zhixin Ma, Cheng Yu-Tong et al. (+3 more)
Cultural understanding in
video means more than recognizing what is visible; it requires grasping the symbolic and temporal significance of cultural concepts. We decompose this into three abilities: naming what a concept symbolizes, visually recognizing it on
video, and locating its sub-events in time. Existing
video-cultural
benchmarks tend to test what is seen, collapsing these three abilities into a single score that hides the bottleneck. We introduce the Cultural Moment
Benchmark (CMB): 306 expert-curated concepts from seven countries in Southeast Asia across five categories. We evaluate each concept through three stages, one per ability. Given a description, Stage 1 (S1) selects from four candidate concept names, Stage 2 (S2) selects from four candidate
video moments, and Stage 3 (S3) predicts the start and end times of the moment in a
video. To keep each stage focused on a distinct ability, we use three design choices: semantic-similarity distractors (S1, S2), unlabeled
video moments (S2), and free-form localization on a different example
video (S3). Across six
vision-
language models, failure modes vary by ability and modality. i) Even the strongest closed-source models score below 30% when all three stages must be correct; ii) The three abilities do not fully cascade: naming a concept correctly helps half the models recognize it on
video, but recognizing it has little effect on locating the sub-event in time; iii) Audio is complementary, redundant, or distracting depending on the concept, more often distracting in non-Latin-script countries; removing both audio and subtitles hurts Games and Music the most. Our 14-rater human study shows that even Expert raters score below chance on concepts from a neighboring country, indicating that CMB requires country-specific cultural knowledge. CMB acts as a diagnostic harness, attributing failures to a specific ability or modality.
Read the full abstract ▾
All authors Burak Satar, Zhixin Ma, Cheng Yu-Tong, Huy Hoang Tran, Phuong Anh Nguyen, Chong-Wah Ngo
Information Theory
Compressed sensing matrices from orthogonal spaces over finite fields of odd characteristic
Kanittakorn Moonchaisook, Poom Kumam and Songpon Sriwongsa
In this paper, we construct deterministic
matrices from subspaces of orthogonal spaces over finite fields of odd characteristic and investigate their applicability to compressed sensing. The construction is based on incidence relations among three types of subspaces, yielding families of
matrices with explicitly computable dimensions and coherence. Using coherence-based estimates, we establish sufficient conditions under which these
matrices satisfy the Restricted Isometry Property for prescribed sparsity levels. We also provide numerical comparisons with DeVore's deterministic construction to illustrate the trade-off between the number of measurements, coherence, and sparse recovery guarantees.
Read the full abstract ▾
Artificial Intelligence
Improving O-RADS Risk Stratification from Ultrasound Reports: A Comparative Evaluation of Hybrid versus End-to-End LLM Reasoning Strategies
Xiaotong Tan, Chunli Qiu, Xin Liu et al. (+11 more)
Background: Automating clinical guideline-based decision-making with large
language models (LLMs) remains challenging because of reliability,
hallucination, and limited interpretability. We compared the performance of LLMs and
reasoning strategies for automated Ovarian-Adnexal Reporting and Data System (O-RADS) classification from free-text pelvic ultrasound reports. Methods: In this retrospective study, consecutive patients with ovarian masses who underwent pelvic ultrasound were included. Eight LLMs were tested with three
reasoning strategies: implicit-knowledge end-to-end, rule-informed end-to-end, and a feature-based hybrid architecture that decoupled feature extraction from rule-based classification. The reference standard was O-RADS categorization established by expert consensus. Results: A total of 310 women with 390 ovarian masses were evaluated. The feature-based hybrid architecture using Gemini 3.6 Flash demonstrated the best performance, achieving an accuracy of 99.2% (387 of 390) and almost perfect agreement with the reference standard (weighted kappa = 1.00; 95% CI: 0.99-1.00). Its performance surpassed that of original clinical reports (accuracy, 87.7% [342 of 390]; weighted kappa = 0.94; 95% CI: 0.91-0.96) and end-to-end LLM strategies (accuracy range, 65.6% [256 of 390] to 95.9% [374 of 390]). For structured feature extraction, Gemini 3.6 Flash demonstrated higher overall accuracy than Claude Fable 5 (98.9% vs 97.8%; P < 0.001). The hybrid architecture reduced misclassification errors and mitigated the overstaging tendency observed in original reports. Conclusion: The feature-based hybrid LLM architecture that separates clinical feature extraction from deterministic guideline execution enables highly accurate, reliable, and interpretable automated O-RADS classification, providing a promising approach for standardized, guideline-based clinical decision-making.
Read the full abstract ▾
All authors Xiaotong Tan, Chunli Qiu, Xin Liu, Qing Huang, Guangli Zhou, Bo Gao, Xiaoyan Song, Shuyan Wang, Xiuqin Wang, Wufeng Xue, Ruobing Huang, Dong Ni, Guowei Tao, Jun Cheng
Social and Information Networks
Atomic Common-Day Invoice Clearing under Causal Daily Scheduling: Path-Enabled and Bounded-Cycle Policies
Peplluis Esteva de la Rosa and Sai Srikanth Madugula
Late payment propagates working-capital pressure through supply networks because firms are simultaneously creditors and debtors. We develop an atomic-record temporal invoice-graph method for path-enabled clearing and compares it with complete-candidate bounded-cycle netting under a causal daily greedy schedule. Each invoice remains a residual record with its issue date, due date, amount, and identifier. A candidate is executable through source capacity active on every supporting edge on one common day. A non-bilateral two-edge path reduces two invoice legs, creates a direct settlement instruction between the endpoints, and preserves net positions for all participants on the combined invoice-plus-instruction state; payable-mass reduction is distinguished from invoice compression. The empirical sequence contains 749,952 invoices issued from 2012 through 2023, totalling EUR99.7 billion, and maintains one rolling state across annual boundaries so bridge invoices are introduced once and residual balances continue. Path clearing reduces 48.2% of issue-cohort mass, compared with 43.3% for length-eight cycle netting, an advantage of EUR4.841 billion and 4.855 percentage points. It leads materially in ten cohorts, is practically tied in 2013, and trails in 2012. The result survives alternative ordering, mixed path-cycle policies, cycle-length sensitivity, component resampling, acceleration constraints, and fragment replay. In reciprocal 2022, where almost all local paths lie in the cyclic core, path clearing still leads. Tractable full-information linear programs show path advantage and path-cycle complementarity. The findings establish a policy-level benefit, not global optimality or welfare dominance,and motivate future asynchronous
agent-to-
agent clearing built around deterministic common-day verification,private mandates,reservations, explicit consent,and atomic commit.
Read the full abstract ▾
Artificial Intelligence
LLM-based Agents for Forecasting and Prediction: Methods, Training, Evaluation, and Applications
Xiaogang Xu, Jiaqi Tang, Jianmin Chen et al. (+12 more)
Large
language models (LLMs) now support
forecasting systems that combine
language-based
reasoning with temporal data, evidence
retrieval, external tools, and iterative prediction. We investigate LLM-based
forecasting agents, meaning systems in which a
language model contributes to a scored prediction about a future or currently unobserved target. We organize architectures into three groups. Standalone LLM workflows operate on encoded time series or event context. Tool- and
retrieval-augmented
agents incorporate external evidence. Hybrid systems pair LLMs with statistical or foundation models. We then review training methods and evaluation protocols. We examine negative as well as positive evidence, including sensitivity to small input perturbations, ablations in which the LLM component does not improve accuracy, and
benchmark gains that may reflect contamination instead of temporal
reasoning. We cover applications in finance, weather, health, energy, and operations, and we summarize the
benchmarks and
datasets used for evaluation. The evidence indicates that measurement is a central limitation. Future work requires calibration under distribution shift, contamination-resistant live evaluation, explicit reporting of cost and accuracy together, and methods for handling feedback between deployed forecasts and the outcomes being forecast.
Read the full abstract ▾
All authors Xiaogang Xu, Jiaqi Tang, Jianmin Chen, Yingying Yan, Zhenchao Tang, Xiangxin Zhou, Xiaobin Hu, Wei Wei, Jinfeng Wu, Qifeng Chen, Lu Zhou, Jiafei Wu, Zhe Liu, Jianwei Yin, Weimin Zheng
Machine Learning
Macro-Action Topological Navigation under Noisy Localization using Reinforcement Learning
Simon Hakenes and Tobias Glasmachers
Navigating large, photorealistic 3D apartments from raw pixels is widely considered infeasible for plain reinforcement learning. We build an
agent that does it anyway, estimating its own pose from the camera alone. The
agent has to reach several target objects in sequence, and their positions change between episodes, so it must explore to find them. It builds on our earlier object-centric topological controller, which still read the
agent's true pose and its object detections from the simulator. Here we replace that true pose with an onboard, object-centric estimate. For each object we keep a bank of ORB features that, when the object is seen again, yield a rough pose measurement, which a minimal Extended Kalman Filter (EKF) fuses with a motion model. As on a real
robot, the executed motions are noisy. The estimate drifts, but the
agent and the nearby objects drift together, so a locally consistent pose is enough to follow each short edge and then home in visually on the target, which lets us replace full SLAM with a much smaller model, closer to how biological
navigation appears to work. In the photorealistic Habitat simulator, the
agent reaches its target objects from
vision alone, with a pose that only needs to be locally consistent.
Read the full abstract ▾
Machine Learning
Graph Representation Learning of Lightweight IoT Ciphers
Jonathan Cook, Sabih ur Rehman and M. Arif Khan
SIMON and SIMECK belong to a family of Lightweight Cryptographic
Algorithms (LCAs) based on the Feistel block cipher, designed for Internet of Things (IoT) devices. As with all Feistel ciphers, they are susceptible to differential cryptanalysis, necessitating rigorous resilience evaluations. While state-of-the-art techniques leverage heuristics and sampling to improve efficiency, little work has applied Machine Learning (ML) guided Graph Representation Learning (GRL) to efficiently identify and visualise high-probability differential clusters. We address this gap by introducing an efficient feature engineering strategy that extracts four differential attributes from a partial Difference Distribution Table (pDDT), revealing structural information concealed in raw differential data. Utilising the enriched features, we construct and compare three ML-guided directed graphs for SIMON$32$ and SIMECK$32$ using K-Nearest Neighbour (KNN), Decision Trees (DT), and Random Forests (RF). To the best of our knowledge, our framework produces the first graph-based visualisation of the differential clustering effect, in which high-probability single-bit differentials form geometrically close clusters in the learned
embedding. All three models achieve a precision of $1.0$ in identifying high-probability differentials, confirming zero false positives. KNN achieves the strongest cluster separation, the highest F1 score and the lowest graph construction time of approximately $2.3$ seconds, while DT and RF produce optimal paths with near-perfect regression. The results are consistent across both LCAs, demonstrating the applicability of the framework to other AND-rotation LCA families.
Read the full abstract ▾
Human-Computer Interaction
What Makes an Initial Reaction Ready for Discussion?: Multi-Persona AI Support for Stance Reflection and Writing
Sky Shih-Kai Hong, Mu-Tien Kuo, Wei-Ji Chen et al. (+1 more)
An initial reaction to a social or community issue can feel meaningful before it is ready to become a message: people still need to clarify the claim, anticipate audience risks, and decide how much
reasoning should become visible to others. We present StanceLab, a prototype for preparing a stance before entering a discussion. The prototype compares a three-persona mode, where an Interviewer, Mentor, and Opponent respond in parallel to help users diagnose and revise a stance, with a standalone LLM mode. In a formative within-subject pilot with six participants and 12 task sessions, every session produced a short final message in the notepad. The pilot revealed two design requirements: persona roles should diagnose useful blind spots or objections, and parallel responses need coordination support. We propose a future diagnosis-and-writing workflow that turns persona-based reflection into selective, audience-aware final messages.
Read the full abstract ▾
All authors Sky Shih-Kai Hong, Mu-Tien Kuo, Wei-Ji Chen, Dennis Wang
Machine Learning
Reservoir of Importance: Learning Semi-Structured Sparsity with Differentiable Subset Sampling
Ha Dinh, Xuan Duy Ta, Khoat Than et al. (+1 more)
Semi-structured $N$:$M$ sparsity has emerged as a practical direction for accelerating large
language models (LLMs). However, existing learnable-mask approaches incur substantial parameter and
memory overhead, limiting their scalability to large models and aggressive sparsity regimes. In this work, we revisit semi-structured
pruning from a perspective that reconciles efficiency with scalability. We propose Reservoir of Importance (RoI), a lightweight semi-structured
pruning framework that learns sparsity masks through differentiable subset sampling. Unlike prior methods that model full categorical distributions over all feasible $N$:$M$ patterns, RoI introduces a compact-logit parameterization for sparsity mask learning and performs sampling without replacement to select masks, thereby reducing trainable parameters from combinatorial complexity to $\mathcal{O}({M})$. As a result, RoI requires 1.5-8.75$\times$ fewer learnable parameters and significantly lower
memory cost, while remaining fully aligned with hardware-friendly sparsity patterns. Extensive evaluations across multiple scales of the Qwen2.5 LLM family (0.5-7B parameters) demonstrate that RoI achieves competitive performance with strong
memory efficiency, stability, and scalability to more aggressive $N$:$M$ sparsity patterns, offering a practical path toward efficient LLM deployment.
Read the full abstract ▾
All authors Ha Dinh, Xuan Duy Ta, Khoat Than, Khac-Hoai Nam Bui
Computation and Language
Beyond Verdicts: A Graph-Based Analysis of Human and LLM Reasoning in Scientific Fact-Checking
Abdul Ghafoor, Muhammad Arslan Manzoor and Yufang Hou
Misinformation that cites legitimate papers can be especially harmful when it distorts what those studies actually report. While existing automatic fact-checking systems based on large
language models (LLMs) can assess whether a model assigns an Incorrect verdict and can gen- erate explanations for that decision, they typi- cally do not indicate whether the model follows the same
reasoning path as human experts or arrives at the verdict through a different but still valid path. In this work, we introduce a graph- based framework (typed
reasoning graph) for comparing human and LLM
reasoning paths in scientific fact-checking. Building on prior work on fallacious
reasoning in biomedical misinformation, MISSCIPLUS (Glockner et al., 2025), we model each explanation as a rea- soning graph that links the false claim to the relevant study context, study findings, fallacy- supporting premises, and fallacy labels. This representation enables one-to-one
alignment of human and LLM
reasoning at the level of fallacy-specific sub-graphs. For non-human- aligned LLM paths, we validate grounding in the cited study, relevance to the claim, and suf- ficiency for the verdict. Using 84 false claims from MISSCIPLUS, we evaluate GPT-5, Claude Opus 4.7, and Qwen3-32B across
prompt and evidence settings. Results show distinct perfor- mance dimensions: Qwen3-32B has the lowest verdict failure rate, GPT-5 the highest human
alignment, and Claude Opus 4.7 weak verdict prediction but often valid
reasoning in success- ful cases
Read the full abstract ▾
Artificial Intelligence
From Inertia to Objectivity: Improving Deep Research Agents with Noise Isolation
Xiangxin Zhang, Zhanwei Zhang, Zhihang Fu et al. (+2 more)
Web search
agents powered by Large
Language Models (LLMs) show strong promise, but deep research tasks expose a recurring failure mode: once an
agent has produced a query, plan, or intermediate conclusion, it becomes less objective when later judging the consequences of that same action. We term this phenomenon inertia bias. To make it measurable, we introduce the IBIS
benchmark, which controls the search observations while varying whether the model is evaluating the outcome of its own prior action. We find that models are substantially worse when they "own" the preceding search step, showing that self-authored action history can systematically distort subsequent judgment. We further show that this bias propagates into two forms of system-level degradation: search noise at the worker level and contextual noise at the manager level. To address this problem, we propose NIS-
Agent, which applies context isolation at the two decision points most vulnerable to inertia bias: webpage triage and final-answer validation. Across
GAIA, WebWalkerQA, BrowseComp, and BrowseComp-zh, NIS-
Agent achieves competitive performance while reducing
token cost by 33% compared to our baseline. We further train an 8B model to be intrinsically more resistant to inertia bias; under the same NIS-
Agent framework, it attains average performance comparable to GPT-4o on deep research
benchmarks. Our code is publicly available at https://github.com/PangSMPang/NIS-
Agent.
Read the full abstract ▾
All authors Xiangxin Zhang, Zhanwei Zhang, Zhihang Fu, Binbin Lin, Wenxiao Wang
Distributed Computing
Introducing the Self-Stabilizing SLEEPING Model
Tistou Fages, Colette Johnen and Mikaël Rabie
The SLEEPING LOCAL model introduces a new complexity parameter, the awake complexity, to make distributed
algorithms energy-efficient. In the synchronous LOCAL model, nodes can now decide to be awake or asleep in each round. In a round, only awake nodes can communicate to share information, which consumes energy. The awake complexity is the maximum number of times a node is activated to produce an output. In particular, it often comes at the cost of the total number of rounds required to solve a problem, compared with
algorithms in which every node is awake in every round. In this article, we adapt the notion of awaken rounds to the context of self-stabilization, introducing the Self-Stabilizing SLEEPING model. Nodes are no longer required to remain awake at all times. However, in self-stabilization, nodes must be activated infinitely often to detect any issue in the system's current state. In this model, the complexities are: * How many synchronous rounds are needed to reach a legitimate configuration? * How many times does a node need to be awake to reach this configuration? * How often does a node need to be awake once this configuration is reached? The goal is to minimize those three metrics, and we can expect different trade-offs. We present energy-efficient
algorithms to solve the problems of finding a $(\Delta+1)$-coloring, a Maximal Independent Set, and a Maximal Matching, thanks to new ad hoc sleeping techniques that reduce the awake complexity (i.e., energy consumption) during the convergence phase. We also propose two
transformers that adapt silent self-stabilizing
algorithms to the SLEEPING setup. The first
transformer is pretty simple and deals with low-complexity
algorithms. The second is more elaborate and is more energy-efficient when it transforms slow self-stabilizing
algorithms.
Read the full abstract ▾
Artificial Intelligence
AutoSaddler: Automatic Harness Optimization with Durable Updates from Agent Execution Traces
Sungho Park, Wonjoong Kim, Rongyuan Tan et al. (+10 more)
LLM
agents remain unreliable on long-horizon tasks, where small local failures can compound over extended interactions and lead to overall task failure. Although external harnesses can substantially improve robustness, harness design remains a manual and expensive process that requires searching over a large space of
prompts, tool configurations, and control logic. We propose AutoSaddler, an automatic harness optimization framework that formulates harness improvement as an offline learning problem and iteratively updates the harness using failure signals from mini-batches. AutoSaddler combines failure-trace diagnosis, structured patch generation that treats the harness as code, and validation-based update selection. Experiments on
GAIA2, SWE-Bench Pro, and Terminal-Bench 2.0 show that AutoSaddler substantially improves
agent performance over the corresponding base harnesses, achieving gains of 9.0, 9.6, and 10.0 percentage points, respectively. Ablation studies further suggest that effective harness optimization benefits from three ingredients: deep debugging rather than shallow reflection, targeted modifications rather than unconstrained editing, and generalization-aware selection rather than trajectory-specific repair. Together, these results suggest that automatic harness optimization is a promising path toward more performant and reliable
agent systems.
Read the full abstract ▾
All authors Sungho Park, Wonjoong Kim, Rongyuan Tan, Jue Zhang, Wook-Shin Han, Pengfei Gao, Chanyoung Park, Yongqiang Yao, Rao Fu, Elsie Nallipogu, Qingwei Lin, Saravan Rajmohan, Dongmei Zhang
Robotics
RoboRacer Arena: Specification-Driven Track Construction for Autonomous Racing
Mihaela-Larisa Clement, Agnes Poks and Ezio Bartocci
Learning-based autonomous racing relies on diverse training environments, yet constructing a new track requires
geometric design, validation, and simulation-asset generation. This coupling makes track geometry difficult to vary systematically during policy training and evaluation. To address this limitation, we present RoboRacer Arena, a specification-driven pipeline that exposes track geometry as an explicit experimental variable. Starting from natural-
language requirements, a seeded coverage-guided constructor generates closed-loop layouts, validates the exported occupancy maps, and automatically builds the corresponding Isaac Sim environments. The same interface admits recorded maps and scaled
circuits, yielding an initial reference collection of 130 tracks. Across our evaluation, RoboRacer Arena achieves the highest valid-map generation rate among the tested construction procedures under their respective computational budgets and converts eight
benchmark maps into simulation assets in less than 2.5 seconds each. We further reconstruct the RoboRacer vehicle as a CAD and USD asset and use it for parallel residual-policy training and deployment on the physical platform. In physical experiments, policies trained with RoboRacer Arena complete ten consecutive laps at command settings up to four times the nominal training speed. These results demonstrate that explicit track requirements can be connected to validated simulation assets and physical evaluation within a reproducible autonomous-racing workflow.
Read the full abstract ▾
Digital Libraries
Beyond FAIR Data: Instrument Traces for Active and Autonomous Scientific Experimentation
Sergei V. Kalinin, Boris N. Slautin, Yu Liu et al. (+1 more)
Artificial intelligence is turning scientific instruments into active systems in which observations can determine what is measured next. We argue that this creates an additional scientific record, the experimental trajectory, complementing sample provenance, acquired data and metadata, and analysis workflows. Instrument Traces should ultimately be synchronized with Sample Traces describing specimen
evolution and Decision Traces recording human or algorithmic choices. We reconstruct an Instrument Trace retrospectively from a longitudinal AFM/PFM archive containing 118,000 timestamped events from 2023-2026. Conventional saved files reveal material campaigns, latent probe and calibration states, session-level complexity, experimental decision grammar, and composite tuning actions. They also expose what is missing, including unsaved tuning and failures, explicit sample/probe identities, complete timing, exogenous state, and decision rationale. We therefore propose a prospective trace architecture that records synchronized sample, instrument, and decision histories, enabling reproducible autonomy, predictive maintenance, counterfactual analysis, operator training, and transfer across facilities.
Read the full abstract ▾
All authors Sergei V. Kalinin, Boris N. Slautin, Yu Liu, Charles Cao
Sound
LipsAM: Lipschitz-continuous Neural Networks for Convergent Plug-and-Play Audio Signal Recovery
Kazuki Matsumoto, Ren Uchida, Natsuki Yoshino et al. (+1 more)
The Lipschitz continuity of deep neural networks (DNNs) is essential for establishing theoretical guarantees regarding their behavior. From both theoretical and practical perspectives, various methods have been proposed to construct Lipschitz-continuous architectures and control their Lipschitz constants. However, several DNN architectures common in audio signal processing fall outside the scope of existing theoretical frameworks, hindering the development of Lipschitz-continuous models in acoustic applications. In particular, despite their widespread adoption, DNNs that separately process the magnitude and phase of complex-valued signals cannot be Lipschitz continuous under existing frameworks. In this paper, to address this limitation, we establish a theoretical foundation for constructing amplitude modifiers (AMs), a class of DNN architectures that operate solely on the magnitude of a complex-valued input, with provable Lipschitz continuity. Specifically, we derive a necessary and sufficient condition for an AM to be Lipschitz continuous and propose LipsAMs (Lipschitz-continuous AMs) corresponding to common architectures for audio signals, including time-frequency masking. Furthermore, we develop an efficient framework for evaluating their Lipschitz constants and analytically derive these constants for some of the proposed architectures. As an application, we propose CoReM-LipsAM (Controlled Residual Maps via LipsAM) for plug-and-play (PnP) audio signal recovery, integrating a DNN as a data-driven prior within a model-based signal processing
algorithm. The convergence of the obtained PnP
algorithm is structurally guaranteed by the CoReM-LipsAM architecture and empirically validated through
speech dereverberation experiments.
Read the full abstract ▾
All authors Kazuki Matsumoto, Ren Uchida, Natsuki Yoshino, Kohei Yatabe
Computation and Language
Beatrice Fiumanò, Nicolas Lazzari, Simone Paolo Ponzetto et al. (+1 more)
This paper introduces the
Multilingual FrameNet Corpus (mFNC), a novel resource that extends the English Berkeley FrameNet corpus by collecting and harmonizing existing
language-specific corpora across nine additional
languages: Brazilian Portuguese, Chinese, Dutch, French, German, Italian, Korean, Latvian and Swedish. By training models that rely on different architectures on the mFNC, we consistently outperform existing state-of-the-art Frame Semantic Parsers in both
multilingual and cross-lingual settings, underscoring the importance of
multilingual training data. The mFNC and our trained FSP models are openly available at https://github.com/beatrice-f/mFNC.
Read the full abstract ▾
All authors Beatrice Fiumanò, Nicolas Lazzari, Simone Paolo Ponzetto, Valentina Presutti
Artificial Intelligence
MobilePA-Bench: Benchmarking Mobile Planner Agents on Complex Real-World Tasks
Yi Zhu, Xiongwei Wu, Qiyi Wang et al. (+8 more)
Mobile planning requires a central
agent to coordinate tools, persistent user context, reusable procedures, and specialized sub-
agents while adapting to execution feedback. Evaluating these workflows requires distinguishing task completion from compliance with prescribed resource and procedural requirements. We introduce MobilePA-Bench, a
benchmark comprising 1,705 tasks across eight query task domains and 212 mobile tools in a stateful simulated environment. It combines three features: (1) unified capability coverage, profiling Tool Use,
Memory Usage, Skill Usage, and Sub-
agent Collaboration under a common mobile protocol; (2) evidence-aligned verification, matching each task to a fixed tool-call, state-change, or interaction-behavior checker rather than imposing one completion criterion on all tasks; and (3) capability-gated evaluation, requiring task completion alongside prescribed
memory checks and, in skill-only routing, complete route loading before business-tool use. With MobilePA-Bench, we systematically evaluate open- and closed-weight LLMs, revealing capability-specific weaknesses in complex mobile task execution. Fine-grained analyses distinguish failures in resource
retrieval and procedural compliance from downstream execution errors, providing diagnostic insights beyond aggregate task-success rates. We further train a Qwen3.8-27B baseline with online reinforcement learning in the sandbox, whose improved performance provides initial evidence of the environment's utility for planner development.
Read the full abstract ▾
All authors Yi Zhu, Xiongwei Wu, Qiyi Wang, Tingyu Qu, Jiajun Liu, Sihan Cao, Long Chen, Weigao Sun, Feida Zhu, Yiran Zhong, Steven Hoi
Machine Learning
ST$^2$U: Stateful Test-Time Unlearning via Restricted Knowledge Boundary Control
Xunlei Chen, Qinghui Gong, Ruini Xue et al. (+3 more)
Controlling restricted knowledge in large
language models is essential for model
alignment and safe deployment. Test-time unlearning avoids costly retraining and parameter updates by intervening only during inference. However, existing activation-editing methods apply isolated pointwise corrections, overlooking how autoregressive generation continually reconstructs hidden states from the
prompt, cache, and generated prefix. Consequently, later states may return to restricted knowledge regions after a locally successful correction, causing restricted knowledge re-entry. In this work, we propose Stateful Test-Time Unlearning via restricted knowledge boundary control (ST$^2$U), which formulates test-time unlearning as trajectory-wide boundary control. ST$^2$U first models restricted knowledge boundaries in low-dimensional invertible coordinates while leaving orthogonal non-target components unchanged. During inference, ST$^2$U monitors risk along the trajectory, applies minimal boundary corrections with contextual anchoring, and propagates historical correction states across
tokens to mitigate knowledge re-entry. This trajectory-wide control enables more persistent forgetting while preserving non-target capabilities and limiting inference overhead. Across three
benchmarks and three model families, ST$^2$U delivers the strongest overall balance, combining best or second-best retention with competitive forgetting and substantially less restricted knowledge re-entry than test-time baselines (13.76–19.84% versus 30.44–59.10%).
Read the full abstract ▾
All authors Xunlei Chen, Qinghui Gong, Ruini Xue, Yaodong Hu, Tian Lan, Wenhong Tian
Machine Learning
FedCC: Towards Addressing Label Distribution Skews in Distillation-Based Federated Learning
Wenxuan Ye, Onur Ayan, Xueli An et al. (+1 more)
Federated Learning (FL) enables distributed clients to collaboratively train models without sharing raw data, making it promising for leveraging massive devices in communication networks. In
distillation-based FL, each client applies its local model on an unlabeled public
dataset, and shares only prediction results with the server. While heterogeneous local data introduces label distribution skew, thus biasing client models toward majority classes and leading to potentially inaccurate predictions. The lack of ground-truth labels in the public
dataset hampers the server's ability to calibrate predictions, which ultimately degrades overall performance. To address this, we propose FedCC, a simple and effective
algorithm for mitigating client misclassification. Instead of being forced to classify and risking error propagation, clients are allowed to tag ambiguous samples as 'unknown'. This additional class, together with calibrated pseudo-labels on the public data, balances confidence in majority classes against uncertainty in under-represented ones. Extensive experiments demonstrate that FedCC significantly outperforms existing methods, especially under severe label skew. In the extreme scenario where each client holds samples from only one of ten classes, FedCC achieves 67.3% accuracy, while baselines collapse to near-random results.
Read the full abstract ▾
All authors Wenxuan Ye, Onur Ayan, Xueli An, Georg Carle
Artificial Intelligence
Artificial Empathy: Towards a Framework for Unsupervised Agency Detection and Policy Reconstruction
Peter Kuhn, Chris Pang and Sonakshi Chauhan
We study how an AI system can identify and model other
agents in its environment from observation alone, which is a capability necessary for cooperative behaviour in the real world. This problem is less constrained than inverse reinforcement learning and remains largely unexplored. We propose a framework that uses a reinforcement learning
agent, trained on an independent task as a prior about agentic dynamics, to perform agency detection and policy reconstruction.
Read the full abstract ▾
Computation and Language
Meta-Moderator: Empowering Multi-Agent Debate with Meta-Cognition
Wentao Hu, Zhuoyue Wan, Jinhao Shen et al. (+3 more)
Multi-
agent debate can improve large
language model
reasoning by eliciting diverse hypotheses and critiques, yet its performance is often constrained by weak moderation. Common pipelines rely on fixed budgets, agreement-based stopping, or untrained judges, leading to redundant deliberation and unreliable evidence aggregation. We cast moderation as a meta-cognitive process, monitoring debate utility, controlling deliberation, and adjudicating a final answer, and introduce Meta-Moderator, a learnable framework that dynamically regulates debate and decides when to finalize an answer. Meta-Moderator is trained independently of the debaters via outcome-driven policy optimization, making debate regulation an explicit capability rather than an incidental effect of prompting. Across five
benchmarks, Meta-Moderator outperforms widely used decision layers and transfers across tasks and system configurations. Further analyses show that it allocates debate more selectively and reduces mis-aggregation after informative hypotheses appear.
Read the full abstract ▾
All authors Wentao Hu, Zhuoyue Wan, Jinhao Shen, Chen Jason Zhang, Xiaoyong Wei, Qing Li
Artificial Intelligence
PsychJail: Exploring Psychological Jailbreaks via Multi-Turn Persuasion of LLM Policies
Zeyu Feng, Qingyu Wu, Yuzhe Luo et al. (+1 more)
Large
language models (LLMs) are increasingly deployed in education, healthcare, policy advising, and other interactive settings, where users engage them as sustained social interlocutors rather than one-shot query engines. This shift makes jailbreaks a growing safety threat, yet most research emphasizes single-turn
prompt optimization or iterative attack refinement, leaving psychologically grounded multi-turn vulnerabilities underexplored. We present PsychJail, a psychology-guided framework for red teaming aligned LLMs through theory-grounded, multi-turn persuasion. PsychJail maps established social-psychological persuasion techniques into a tactic-conditioned attack policy. It factorizes each attacker action into a Change-of-Meaning analysis, tactic selection, and victim-visible message, operationalizing the Persuasion Knowledge Model (PKM). The policy is refined with trajectory-level reinforcement learning using a PKM-gated reward that credits early jailbreak success only when every turn contains a well-formed Change-of-Meaning analysis. Across four aligned victim models, PsychJail achieves the highest average attack success rate (87.3%) and outperforms strong single-turn and multi-turn baselines on every model. We also measure susceptibility at the action that breaks each victim, revealing four distinct model-level fingerprints that identify which persuasion levers affect each model and how broadly. These fingerprints help explain cross-model transfer asymmetry. We interpret them as four candidate psychological profiles-rationalist, credibility-driven, narrative-monoculture, and broadly persuadable-while treating this interpretation as a
conjecture requiring future validation. Our findings establish psychological jailbreaks as a distinct red-teaming frontier for increasingly interactive LLMs.
Read the full abstract ▾
All authors Zeyu Feng, Qingyu Wu, Yuzhe Luo, Hua Cheng
Computation and Language
Beyond Surface Cues: Disentangling Sociocultural Signals in Multilingual LLMs
Yuanjun Feng, Tanzhou Liu, Stefan Feuerriegel et al. (+1 more)
Multilingual LLM outputs can vary across sociocultural contexts. However, evidence of cultural grounding can be misleading: identity labels may be inferred from explicit or indirect textual cues, while names and wording can reveal the source
language. Treating all these signals as evidence of cultural grounding may obscure potential biases. We present a human-validated, multi-
agent audit that separates three questions: whether outputs reproduce social biases, whether identity groups are represented differently, and whether outputs reflect cross-cultural patterns. The study analyzes 89,253 outputs from 12 LLMs in English, French, and Chinese, spanning 18 occupations and three task conditions. We find that bias representation varies systematically across
languages and tasks. Removing direct identity cues sharply reduces identity-label prediction in English and Chinese, but has a much smaller effect in French. Across all
language-genre settings, the cultural context associated with the source
language receives the highest average relevance score, with moderate agreement between automated and human ratings. However, the ability to identify the source
language drops substantially after translation and again after masking names. Without these controls,
multilingual audits may mistake surface cues for cultural understanding, leading to misleading conclusions about cross-cultural variation and bias. Our audit offers a practical framework for separating such shortcuts from more meaningful cross-cultural patterns.
Read the full abstract ▾
All authors Yuanjun Feng, Tanzhou Liu, Stefan Feuerriegel, Yash Raj Shrestha
Computer Vision
When the Edit Changes the Patient: Measuring Identity Preservation in Counterfactual Retinal Images
Andrea Posada, Wenke Karbole, Bach Ngoc Doan et al. (+9 more)
Counterfactual medical image generation aims to modify an existing image to reflect a hypothetical scenario in which certain characteristics of the imaged subject are altered, while keeping their identity fixed. Most existing works repurpose established image editing methods, which do not directly supervise identity preservation. Instead, they assume that identity is implicitly preserved by anchoring generation to the source image. This assumption is rarely tested and may fail in domains where biometric cues are subtle, such as retinal optical coherence
tomography (OCT). In this work, we explicitly measure identity preservation for three groups of text-conditioned editing methods - source-anchored, structured-
prompt, and paired-training - using referee classifiers,
embedding alignment scores, and a blind reader study. We find that all methods produce high-quality OCT images with comparable editing success, yet their identity preservation differs markedly. Source-anchored editing frequently alters the depicted subject, while paired-training preserves it best. We argue that future work on medical counterfactual generation must explicitly measure and report identity preservation alongside image realism and editing success.
Read the full abstract ▾
All authors Andrea Posada, Wenke Karbole, Bach Ngoc Doan, Alexander Weers, Solmaz Abdolrahimzadeh, Maria Patsiamanidi, Kahkashan Haider, Vaishali Khare, Daniel Rueckert, Andrew Lotery, Sobha Sivaprasad, Martin J. Menten
Computation and Language
Most of the LLM Routing Gap Is Task Type
Janghoon Lee
An LLM router picks which model should answer each query. The appeal is that models fail on different questions. Whatever single model is best overall still gets some wrong, and another model in the pool gets many of those right. Getting that choice right every time is the ceiling, and a router is an attempt to approach it. However, recent work reports that routers do not get close. Across 21 routing methods on five
benchmarks, sharply different designs land within a fraction of a point of each other, and all of them stay far below that ceiling. Learned routers often fail to beat simply always calling the strongest model. We ask what those missed questions have in common. We set fourteen models to answer all 294 questions, with 7 task types across 3
languages: Korean, English and Hindi. We ran the whole
matrix twice, changing nothing, but 5.37% of the 4,116 model-question pairs came out scored differently anyway. Run-to-run movement like that is normal, and we argue that a small win does not show that routing did anything, ours or anyone else's. Counting an answer correct only when the model got it right in both runs, 29 questions on this
matrix can be improved with routing. Every correct-answer count here is on that rule. Task type accounts for most of them: assigning each task type one model in advance, chosen once and never updated, improves 21 of the 29. Splitting each task type by
language improves 2 more and leaves 6 of 294 unoptimized. That handful is what a learned router would have been built for, and it is smaller than the run-to-run movement above, which is a share of pairs rather than of questions. The static table we adopted answers 262 of 294 questions at \$3.33 per run, against the best single model's 245 at \$7.69. All of this is fitted and scored on the same 294 questions with no holdout.
Read the full abstract ▾
Machine Learning · Cross-listed from hep-ph
Neural Boltzmann Equations
Jonas Spinner and Jack D. Shergold
The dynamics of particles in the early universe are described by Boltzmann equations, which involve high-dimensional phase-space integrals. Classical approaches use quadrature integration and evolve the system on a fixed momentum grid, which scales poorly to complicated systems and parameter scans, severely limiting the complexity of processes that can be studied. We introduce Neural Boltzmann Equations (NBEs), which combine three coupled concepts to overcome these limitations. First, particle properties are encoded in physics-inspired neural distribution functions, with parameters that can be predicted using neural networks, enabling efficient parameter scans. Second, phase-space integrals are evaluated with Monte Carlo, using importance sampling tools from
collider physics. Third, we use the natural gradient method to evolve the system. After demonstrating the individual benefits of NBEs, we use the framework to perform the most numerically precise calculation to date of the effective number of relativistic
neutrino degrees of freedom, $N_\mathrm{eff}$, in the Standard Model.
Read the full abstract ▾
Computation and Language
Unlearning Is Not Just Erasing: Temporal Decoupling via Generation Inequality
Xunlei Chen, Qirui Ye, Yuang Li et al. (+5 more)
Large
language models (LLMs) require effective unlearning to address
privacy regulations and safety concerns. However, achieving precise forgetting without compromising general utility remains challenging. Existing sequence- and
token-level methods penalize target outputs without modeling their context-dependent
retrieval paths, which can disrupt linguistic structure or suppress benign knowledge. We present ADU, a fine-grained, training-based framework that shifts unlearning from
token erasure to contextual
attention-pathway decoupling. Exploiting the functional distinction between local and global
attention heads, ADU identifies preplan positions that retrieve persistent sensitive anchors and fixes their candidate paths under the original model. It then trains
attention-projection adapters to suppress
attention mass along these paths while preserving local-
attention structure and retain-set
language modeling. Post-training activation exchange tests whether the modified
attention-output module transmits the learned forgetting effect. ADU achieves the strongest aggregate performance among evaluated baselines on the TOFU and WMDP
benchmarks, including a Forget Quality of (0.93) on TOFU. It preserves 87–98% of model utility (92.9% on average versus 81.9% for baselines) while reducing side effects in benign contexts.
Read the full abstract ▾
All authors Xunlei Chen, Qirui Ye, Yuang Li, Yi Gong, Zhaokun Wang, Wenyi Li, Shiyao Guo, Jinyu Guo