Computation and Language
The "10th Juror": Open-Set Standpoint Screening for Bureaucratic Bias Detection
Yuchen Miao, Zijun Wang, Chang Han et al. (+3 more)
Presupposing the boundaries of bias is itself a form of bias. We study closed-loop bias governance for Dutch government documents, where a system must detect biased
language, ground decisions in legal and contextual evidence, rewrite problematic sentences when intervention is warranted, and verify that the rewrite mitigates harm without distorting meaning. Existing methods face three challenges: (i) discriminative classifiers capture surface regularities but lack normative grounding; (ii) zero-shot LLMs often adopt generic viewpoints and over-flag ambiguous administrative
language; and (iii) fixed taxonomies inherit the Closed-World Assumption, missing emerging local targets. We propose MARS-Gov, a standpoint-aware multi-
agent framework that combines legal
retrieval, open-set target screening, specialized jurors, conservative routing, and rewrite verification. When screening finds an uncovered group, MARS-Gov instantiates a dynamic "10th juror" to deliberate outside the fixed panel. On DGDB, MARS-Gov sets a new SOTA with 0.880 F1, outperforming the strongest zero-shot LLM detector by 20.2 points (29.8% relative) and the best supervised Dutch encoder by 6.8 points, while reducing unnecessary interventions to 2.5%. Leave-One-Category-Out (LOCO) evaluation recovers held-out categories with 85.1% Correct@1 and 93.8% Correct@3.
Read the full abstract ▾
All authors Yuchen Miao, Zijun Wang, Chang Han, Yurui Shi, Mingtai Zhang, Siyang Xu
Computation and Language
Can a System-One LLM Perform Knowledge Tracing When Few or No Learners Are Logged?
Unggi Lee and Haeun Park
Knowledge tracing (KT) models need many logged learners, so a new course or platform starts without a usable model. In LLM-based KT the LLM generates the answer, which we call System-Two; it is either fine-tuned on the target data or reasons and votes over ten samples, which is slow and gives coarse probabilities. We ask whether an off-the-shelf System-One LLM, which returns a probability for a typed question directly in a single pass, can perform KT when few or no learners are logged. On seven
datasets, Jev without any data from the target platform reaches a mean AUC of .706, above the best of 28 deep KT models trained on 8 learners (.689) and above System-Two Thinking-KT on all seven
datasets (.650) at about 1/100 of its API cost. Adding examples and a similar-learner statistic from the logged learners (JevKT) raises this to .722; JevKT stays significantly ahead of deep KT up to 16 learners and ahead on average up to 64, and supervised KT catches up between 64 and 128 learners. Among the readers we tested, the gain is specific to Jev, since three other LLMs queried with the byte-identical typed request through the official System-One adapter fall below it on all seven
datasets, and reader swaps and contamination checks find no evidence that the input format or memorised data explain the gain. For new learners the advantage holds from their first interactions, whereas on unseen items with all learners logged, deep KT remains ahead.
Read the full abstract ▾
Machine Learning
QUILT: Rethinking Sparse-Attention Prefill through Shared Query Execution
Zhenduo Zhao, Qihui Zhou, Mingcong Song et al. (+7 more)
Sparse
attention reduces the cost of long-context
attention, but existing kernels typically process queries independently, repeatedly loading and dequantizing KV entries shared across queries. We observe substantial overlap in the KV entries selected by neighboring queries, creating opportunities for cross-query reuse. We present QUILT, a workload-aware sparse-
attention execution mechanism that jointly processes neighboring queries and reuses shared KV entries to reduce redundant
memory traffic and computation. QUILT introduces Shift-and-Compare Set Decomposition (SCSD), which transforms irregular set operations into regular data-parallel primitives suitable for modern accelerators, and pipelines SCSD with
attention computation to hide its overhead. Cascaded sharing captures reuse hierarchically at multiple granularities. A tile-aware execution strategy balances sharing granularity with hardware tile utilization and selectively removes low-importance query-specific tails to eliminate underutilized tiles. We evaluate QUILT on LongBench using GLM-5.3 and DeepSeek-3.2 under both tensor and sequence parallelism. Compared with the state-of-the-art sparse-
attention kernel, QUILT reduces average kernel latency by up to 55.1% and processed KV data by up to 55.9%, while reducing time-to-first-
token (TTFT) latency by up to 36.8% with negligible accuracy degradation.
Read the full abstract ▾
All authors Zhenduo Zhao, Qihui Zhou, Mingcong Song, Zhiyi Chen, Chuangguan Ye, Fengfan Hou, Zequn Gong, Jing Li, Hongjie Si, Guoping Long
Machine Learning
Machine Learning Optimization for Enhanced OS Fingerprinting
Jae Sung Kim, Spencer Ekeroth and Jeremy Neale
Operating System (OS) Fingerprinting is a technique that can be used to identify a network's operating systems by evaluating network traffic in the form of TCP/IP packets. This research will explore the effectiveness of passively identifying operating systems on the CIC-IDS2017
dataset, a collection of over 47 gigabytes of pcap files with their corresponding operating systems. This research also proposes a new command line interface, OsirisML, which uses nPrint to preprocess the data into tabular data and XGBoost to apply ML to the data to generate, retrain, and test ML models. When packets are split randomly between training and testing, OsirisML models reach an accuracy of 97.66% on a down-sampled subset of the Friday capture and 84.69% on the entire capture. On the entire Monday capture, which contains no attacks, OsirisML reaches an accuracy of 73.83% and an F-1 score of 79.38%.
Read the full abstract ▾
Computation and Language
SFT-as-Context Mitigates Forgetting in Supervised Fine-Tuning
Kenan Tang, Andong Hua, Chengxuan Qian et al. (+2 more)
Supervised fine-tuning (SFT) equips large
language models (LLMs) with specialized capabilities, but often comes at the cost of forgetting the general capabilities of their parent models (i.e., the pretrained models before fine-tuning). This trade-off is especially limiting for queries that require both specialized and general capabilities. We introduce SFT-as-context, a training-free method in which the parent model uses the SFT model's response as context to answer the query. This allows the parent model to acquire fine-tuned capabilities from the SFT response through in-context learning while preserving its own general capabilities. Across 19 parent-SFT model pairs and 11
benchmarks, SFT-as-context remains close to the SFT models on fine-tuned capabilities, with gaps of only 2.2 and 2.1 percentage points on AIME 2024 and LiveCodeBench and 2.0 macro MAE on NutriBench-English, while staying within 2.2 percentage points of the parent models on general capabilities on average. Remarkably, it can solve queries requiring both fine-tuned and general capabilities, even when neither the parent nor SFT model succeeds alone. This approach also extends beyond parent-SFT pairs: responses from a small open-source SFT model can improve a strong closed-source LLM, outperforming either model alone. Furthermore, we use a
Bayesian framework to derive theoretical guarantees that bound the error of SFT-as-context relative to the SFT model on fine-tuned capabilities and to the parent model on general capabilities. In addition, we visualize the
attention weights and find that the parent model attends more to useful SFT responses and less to irrelevant ones, suggesting that selective
attention helps the parent model use the SFT response through in-context learning.
Read the full abstract ▾
All authors Kenan Tang, Andong Hua, Chengxuan Qian, Saket Tiwari, Yao Qin
Signal Processing
SEER: Source-Conditioned Emotion Enhancement via Retrieval for Cochlear-Implant Speech
Hsing-Hang Chou, Yun-Shao Lin, Ching-Chin Sung et al. (+1 more)
Cochlear implants (CIs) restore
speech access but weaken cues needed for vocal
emotion recognition. Prior CI-oriented enhancement requires parallel normal/strong recordings and intensity labels. We propose SEER, a
retrieval-based framework that learns which same-
emotion reference helps each source remain recognizable after CI processing. A source-conditioned retriever learns CI-aware utility from sampled emotional voice conversion outcomes, while uncertainty-guided exploration avoids exhaustive pair evaluation; neither parallel recordings nor intensity labels are required. SEER improves Source macro-F1 at N8 by 7.30 points on RAVDESS and 11.66 points on ESD, with significant ESD gains across N4/N8/N16. Sixteen-listener RAVDESS gains are significant across all conditions. Exhaustive analysis finds an aggregate benefit from stronger references but little effect from matching gender or content.
Read the full abstract ▾
All authors Hsing-Hang Chou, Yun-Shao Lin, Ching-Chin Sung, Chi-Chun Lee
Systems and Control
A User-Triggered UAV Dispatching System for Precise and Timely Mountain Search Missions
Mingyang Wang, Yi Hong, Tungchak Lee et al. (+4 more)
Smartphones carried by hikers routinely record routes, locations, and requests for assistance, and these records can provide precise clues for mountain search. However, they are rarely transformed in a direct and timely way into unmanned aerial vehicle (
UAV) missions that support search and rescue. This paper presents a user-triggered
UAV dispatching system that integrates mobile information collection, ground station mission management, and
UAV search into a single workflow. In our system, a mobile application is designed to continuously track user records and synchronize these records to the cloud database, thereby initiating all subsequent search events and
UAV-dispatch missions. A control platform is developed to retrieve records from the cloud database for operator verification and mission scheduling. A
UAV prototype is implemented to execute the approved mission, follow waypoints, report status, relay phone records when available, and record downward-facing
video for later inspection. Simulations are conducted for
UAV-dispatching missions using a terrain-following flight altitude of $80$~m above ground level and demonstrate successful route generation, ordered waypoint traversal, and return to the launch site for landing of the
UAV dispatching. Local software tests verified timeout detection with a 1s threshold and phone-record transfer from the mobile application through the
UAV to the ground station. Outdoor trials showed that the
UAV could fly autonomously and capture downward-facing imagery in which a person, a nearby footpath, and surrounding vegetation were clearly discernible, providing visual cues for operator assessment and further inspection of the search area.
Read the full abstract ▾
All authors Mingyang Wang, Yi Hong, Tungchak Lee, Ashwin Sundar, Kevin Hung, Qubeijian Wang, Yalin Liu
Artificial Intelligence
GameCommBench: A Unified Benchmark and Type-Aware Evaluation for AI-Generated Game Commentary
Qirui Zheng, Zhengteng Lin, Yunyi Xiao et al. (+7 more)
Game commentary is an open-ended generation task requiring
multimodal perception, strategic
reasoning, and contextual knowledge. Existing AI-Generated Game Commentary (AI-GGC) studies remain fragmented across games, modalities, and evaluation protocols, while overlap-based or holistic evaluators fail to capture the functional heterogeneity of commentary. We introduce GameCommBench, a unified
benchmark spanning board games, sports, and esports, with commentary aligned to heterogeneous game contexts and annotated by commentary type. We further propose Type-Aware Commentary Evaluation (TACE), a structured framework for evaluating different types of commentary. We then validate TACE for reliability and human agreement, and use it to
benchmark representative AI commentators. Results reveal non-uniform capability profiles, with live observation and strategic analysis emerging as major bottlenecks. Together, GameCommBench and TACE provide a diagnostic foundation for comparable and interpretable AI-GGC evaluation.
Read the full abstract ▾
All authors Qirui Zheng, Zhengteng Lin, Yunyi Xiao, Junhao Li, Keyuan Cheng, Xingbo Wang, Yongyi Wang, Lingfeng Li, Yunlong Lu, Wenxin Li
Artificial Intelligence
Balancing Reference Guidance and Free Generation in Trajectory Rollouts for Reasoning RL
Hanyu Wang, Nakul Agarwal, Hossein Nourkhiz Mahjoub et al. (+4 more)
A verified reference solution provides a correct trajectory for training a
reasoning model. Alternatively, a prefix of the reference can guide the model in generating a trajectory of its own. How much reference guidance should we provide? We study this question through prefix continuation, where the model continues from a reference prefix and keeps the resulting trajectory if it passes verification, falling back to the reference otherwise. Since both procedures produce correct trajectories, we compare their distributions with the ideal distribution, the model's own distribution conditioned on successful verification. For one continuation, we derive the KL divergence in closed form, which, up to a bounded term, decreases with the product of the probability of generating a different correct trajectory and the reference surprisal, the negative log probability of the reference suffix given the prefix. Since a longer prefix tends to raise the former but lowers the latter, continuation success alone does not determine the preferred amount of guidance. From this analysis, we learn a prefix selector shared across training questions from continuation outcomes, without estimating success probabilities or additional generation. The resulting Adaptive Reference Guidance (ARG) constructs correct trajectories within a fixed generation budget, and we apply it to all-failure groups in Group Relative Policy Optimization (GRPO). Experiments on Qwen3-4B and Qwen3-8B across five mathematical
reasoning benchmarks show that ARG achieves the highest aggregate pass@12 among the evaluated methods with competitive average sampled accuracy.
Read the full abstract ▾
All authors Hanyu Wang, Nakul Agarwal, Hossein Nourkhiz Mahjoub, Ehsan Moradi Pari, Makoto Fukushima, Jinghui Chen, Vaishnav Tadiparthi
High Energy Astrophysical Phenomena
Improved analysis of time delays between radio wavelengths and X-ray of Sgr A* flares in the shock oscillation model
Ramiz Aktar, Toru Okuda and Chandra B. Singh
We present an improved analysis of the time delays between
radio and X-ray
flares from the supermassive black hole Sagittarius A* (Sgr A*), extending our previous shock oscillation model by incorporating Compton scattering. Our simulations predict delay times of 1-4 h between X-ray and
radio flares, 20-40 min and 0.5-1 h between 334 GHz and 4.5 $\mu$m and 2-8 keV
flares, and 7-20 min and 0.5-1 h between 22-43 GHz and 350 GHz
radio flares, respectively, in good agreement with recent observations. The delay time between the
radio emissions (synchrotron) and X-ray emissions (bremsstrahlung) is interpreted as the transit time for sound waves traveling from the central core near the event horizon to the location of the oscillating shock. Additionally, the delay time observed between different
radio wavelengths corresponds to the transit time of sound waves moving between the frequency-dependent effective radii $R_{\rm eff}(\nu)$, where the optical thickness $\tau_{\nu}(R)$ at frequency $\nu$ equals one on the surface of the
accretion disc. The X-ray emission and the
radio emission at lower frequencies peak later than the
radio emission at higher frequencies. This delay occurs because the acoustic perturbations created by synchrotron emission peak near the event horizon and then expand outward. Conversely, an inverse delay happens when the acoustic waves generated by the upstream shock wave travel toward the event horizon. The predicted delay times from this interpretation qualitatively match the numerical results obtained in simulations.
Read the full abstract ▾
Strongly Correlated Electrons
Probing and Manipulating Quantum Materials with Strong-field Terahertz and Mid-infrared Radiation
S. J. Li, H. Y. Wu, S. J. Zhang et al. (+1 more)
Emergent phenomena in
quantum materials arise from interactions among different degrees of freedom. Identifying these interactions and establishing their roles in ordered states are central problems in condensed-matter physics and provide a basis for controlling material properties. Many relevant low-energy excitations and gaps lie in the terahertz (THz) and mid-infrared (MIR) range, where intense phase-stable fields can both interrogate nonlinear dynamics and drive selected degrees of freedom far from
equilibrium. This review surveys recent progress in probing and manipulating
quantum materials with strong-field THz and MIR radiation, with emphasis on nonlinear electronic dynamics, collective-mode coupling,
symmetry-sensitive responses and field-induced changes in electronic structure and ordered states. Particular
attention is given to developments enabled by the Synergetic Extreme Condition User Facility (SECUF).
Read the full abstract ▾
All authors S. J. Li, H. Y. Wu, S. J. Zhang, Y. T. Li
Quantum Physics
Sub-MHz SWAP spectroscopy of two-level systems in superconducting qubits
Joaquin Minguzzi, John Clai Owens, Oskar Painter et al. (+2 more)
An outstanding challenge in superconducting
quantum circuits is mitigating loss from two-level system (TLS) defects. SWAP spectroscopy with flux-tunable
qubits can detect individual TLSs, but the Rabi chevrons that signal coherent exchange have been observed only for the most strongly coupled defects. Here, we introduce a sequence with fast flux pulses only at state preparation and readout, switched off during the energy-relaxation delay for finer frequency resolution and longer interaction times. We detect Rabi chevrons at swap rates as low as 0.1 MHz, an order of magnitude lower than previously reported. In the dressed-state picture, we fit the coherent swap oscillations to extract the TLS energy-relaxation time. These Rabi chevrons reveal coherent TLSs with a median energy-relaxation time of 10 microseconds, at coupling strengths previously associated with incoherent TLSs. The swap coherence times exceed the
qubit T2* limited by flux noise, indicating that the dressed gap protects them from low-frequency 1/f noise.
Read the full abstract ▾
All authors Joaquin Minguzzi, John Clai Owens, Oskar Painter, Matthew H. Matheny, Loren J. Swenson
Group Theory
Conjugacy problems in higher-dimensional Thompson groups
Ville Salo
We prove that topological conjugacy of elements of the Brin–Thompson group $nV$ is undecidable for every $n\geq2$. From a Turing machine $T$, the reduction produces two elements of $nV$, which are conjugate by an involution in $nV$ if $T$ halts on the empty tape. If $T$ does not halt on the empty tape, the topological dynamics of the two group elements are distinguished by their visits to minimal subsystems of the set of (germ-)aperiodic points. In these groups, and any finitely-generated supergroups inside the homeomorphism group of Cantor space, we also obtain undecidability of the Whitehead problem (the problem of being in the same
automorphism orbit), and undecidability of group-theoretic conjugacy.
Read the full abstract ▾
Artificial Intelligence
When Interfaces Speak: Data-Aware Generative UI Harness for Active Interaction
Xiaolong Li, Xiaohan Xu, Jinyang Li et al. (+5 more)
Most human-
agent interaction today remains text-based. Natural
language can impose cognitive overload, ambiguity, information chaos, and slow input for complex tasks; ephemeral generative UIs can present structured information and guide users toward task completion. We propose GenUI-Harness, a multi-
agent harness pairing a Tool
Agent for information
retrieval and task execution with a GUI Coder
Agent that identifies ambiguities and generates front-end code for structured interfaces. Training the coder with reinforcement learning is challenging: verifiable rewards for interactive UI generation require costly execution, while LLM-as-a-Judge rewards are prone to reward hacking. We address the first challenge with Dynamic UX, a lightweight package for dynamic interaction and reward collection in a single sandbox, and the second with Reward Auditor, a meta-reward mechanism that monitors reward distributions and distills diagnostic patterns into a shared rubric and scoring specification. We introduce UI-TAU Bench, a
benchmark for active human-
agent interaction through generated UI code, built on 10 real-world domain databases constructed from public data sources and based on Tau-Bench tool-use settings, with Lite (300 tasks) and Full (1,000 tasks) splits. GenUI-Harness achieves an average Pass@3 gain of 4.48 percentage points over smolagents on Lite. Training with GenUI-Harness improves a 4B backbone from 9.33% to 58.00% Pass@3, outperforming larger frontier models such as Claude Opus 5 (46.67%). GenUI-Harness also remains robust on ambiguous and non-ambiguous queries. In a reviewer survey comparing communication channels, generated UIs reduce average dialogue rounds from 3.4 to 1.2. These results show that data-aware generative interfaces can support effective task completion and reduce dialogue rounds in evaluated database-backed workflows.
Read the full abstract ▾
All authors Xiaolong Li, Xiaohan Xu, Jinyang Li, Xinnuo Xu, Ge Qu, Nan Huo, Jack Williams, Reynold Cheng
Plasma Physics
The Richtmyer-Meshkov Instability of Thermal, Isotope, and Species Interfaces in a five-moment multi-fluid plasma
K. C. Tapinou, V. Wheatley, D. Bond et al. (+1 more)
The Richtmyer-Meshkov instability (RMI) results from the impulsive acceleration of a density interface where either it or the acceleration is perturbed. Density interfaces may arise due to a change in gas species, isotope, temperature or a combination of these. We computationally investigate the effect of interface type on the
plasma RMI, which is relevant for a range of applications, including inertial confinement fusion. We simulate the
evolution of single-mode perturbed thermal, species and isotope interfaces in an ideal ion-electron
plasma using the multi-fluid
plasma (MFP) model. We find that in the MFP model, the
evolution of different types of interface differ significantly, in contrast to single-fluid models where they behave similarly if the Atwood number is matched. The thermal and species interfaces produce the most severe response to shock acceleration, experiencing the secondary instabilities and enhanced primary mode growth. The isotope interface
evolution is restrained in comparison to the former cases, resembling the response predicted by single-fluid models. The determining factor in the severity of the MFP RMI is the density ratio across the initial interface in the electron fluid, which is unity for an isotope interface. We observe that as the density ratio across the electron interface decreases, so do the magnitudes of the self-generated fields and consequently the severity of the growth amplification. Generally, the
evolution of the RMI with different types of interface becomes more similar as the level of coupling between the ion and electron fluids is increased, characterised by reducing the
plasma non-dimensional skin depth
Read the full abstract ▾
All authors K. C. Tapinou, V. Wheatley, D. Bond, Ingo Jahn
High Energy Physics (Phenomenology)
Quantum Magic and the Strong Coupling Constant
Qiaofeng Liu, Ian Low and Zhewei Yin
The three gauge couplings of the Standard Model (SM) can be parameterized as the fine structure constant $\alpha$, the weak mixing angle $\sin^2\theta_w$, and the strong coupling $\alpha_s$. They are fundamental constants whose values remain unexplained. Previously we showed that minimizing magic production in charged-
lepton scattering reproduces $\sin^2\theta_w$ with great precision. Here we study magic production in the six
flavor-diagonal color-singlet channels $q\bar q\to q\bar q$ at tree level, including
photon,
gluon, $W/Z$, and
Higgs exchange. Varying only $\alpha_s$, with all other inputs fixed at each center-of-mass energy, we find that the finite-$\alpha_s$ magic minimum in the top channel reproduces the $\overline{\mathrm{MS}}$ value of $\alpha_s$ to within $5\%$ from $1~$TeV to $10^9~$GeV.
Higgs exchange involving the large top Yukawa coupling is essential for this agreement. All six channels produce near-minimal magic at the SM couplings, and reducing $\alpha_s$ toward $\alpha$ increases their magic production, potentially explaining why the strong interaction is “strong.” To understand why minimizing magic reproduces both $\sin^2\theta_w$ and $\alpha_s$, we
conjecture that the parameters of fundamental interactions among color singlets reflect a principle of
quantum computational efficiency, with the physical universe emerging from a
quantum simulation subject to constrained resources.
Read the full abstract ▾
Numerical Analysis
Numerical Neural Operator: Connection Between Numerical Analysis and the Operator Learning and Building Discretized Neural Operators from Finite Element Methods
Zecheng Zhang
Operator learning aims to learn mappings between function spaces and has been widely used for PDE-related problems. This raises two fundamental questions. First, is there a connection between standard numerical methods and neural operators, given that both are designed to approximate PDE solution operators? Second, when a neural operator is trained on numerically generated data, does it approximate the underlying infinite dimension operator, the training data themselves, or the discretized numerical scheme used to generate them? A key hypothesis in the paper is that the trained neural operator may mimic the discretized numerical schemes that generate the training data. The hypothesis has been partially and preliminary verified by the linear approximation principle proposed in Chen Chen 1995 paper and the following scaling law papers. However, neural operator approximation from the perspective of numerical discretization remains unexplored. In this work, we consider a class of PDEs and derive neural operator architectures from the numerical schemes. Firstly, the numerical basis functions, such as finite element basis functions, can be approximated or directly constructed by learnable neural networks. Then the finite difference schemes for temporal discretization can be unrolled into iterative neural architectures that generate the coefficients of the basis of the resulting neural operator. By leveraging numerical analysis, we establish convergence error estimates for these numerically derived neural operators. Notably, the network size depends
polynomial on the approximation error in $L_2$, improving the nested exponential dependence derived in the general neural operator approximation and aligning with the numerical observations. Finally, numerical experiments demonstrate reduced errors and improved parameter efficiency.
Read the full abstract ▾
Robotics
FOCUS: From Privileged States to RGB-D with Controlled Modality Switching and Representation Alignment
Filip Grigorov, Kourosh Darvish and Nandita Vijaykumar
Vision-based reinforcement learning for robotic manipulation is sample-inefficient because RGB-D observations are high-dimensional and noisy. Privileged state information available in simulation can accelerate training, but its absence at test time creates a train-test modality gap. We propose FOCUS, a single-stage PPO framework that trains the critic on privileged state while automatically regulating whether the actor collects rollouts from RGB-D or privileged-state latents. Regulation is driven by the KL divergence between the action distributions induced by the two modalities, while representation
alignment encourages consistent action selection across them. Together, these mechanisms limit RGB-D rollouts when the actor's action distributions from RGB-D and privileged-state latents disagree. As they align, RGB-D exposure increases, shifting on-policy training toward the RGB-D inputs used at test time. Across five manipulation tasks, FOCUS raises average test success from 0.71 to 0.93 relative to the strongest RGB-D-at-test baseline on each task. When accounting for each method's complete training pipeline, budget-normalized training-success AUC increases from 0.47 to 0.65. On Pick-and-Place, test success rises from 0.47 to 0.86, while AUC increases from 0.12 to 0.61, a 5.0x improvement in learning efficiency over the fixed interaction budget.
Read the full abstract ▾
Artificial Intelligence
OpenProblemBench: Benchmarking AI on Open Problems in the Foundational Theoretical Sciences
Zhiyi Li, Sihan Hu, Tianning Xiao et al. (+4 more)
The next frontier for artificial general intelligence is tackling unresolved scientific problems, calling for
benchmarks that assess progress beyond established knowledge. We introduce OpenProblemBench, a
benchmark of 82 unresolved problems drawn from the mathematics and theoretical physics literature. Each problem supplies the research context, assumptions, and prior progress needed to investigate the question. We select problems whose proposed solutions admit comparatively clear checks of their decisive mathematical or computational claims. Four evaluator models independently assess the correctness, completeness, and degree of progress of each submission without reference solutions. Across seven evaluated configurations, GPT-6-Astra achieves the highest mean judged solve rate of 14.0%, compared with 5.5-6.7% for the evaluated full-size open models and 2.4-3.7% for Flash models. Case comparisons connect stronger outcomes to changes in problem representation, general arguments that extend beyond finite evidence, and proofs of the steps needed to complete a solution. By grounding evaluation in questions arising from the research literature, OpenProblemBench provides a setting for investigating the capabilities and limitations of AI as a contributor to foundational theoretical science.
Read the full abstract ▾
All authors Zhiyi Li, Sihan Hu, Tianning Xiao, Xiansheng Cai, Xiaojun Tan, Youjin Deng, Kun Chen
Computer Vision
MCL: Meta Convolution Layer
Naim Reza, Md Al Amin and Ho Yub Jung
Dynamic convolution enhances convolutional neural networks (CNNs) by adapting kernels to input content, but it expresses the effective kernel as a linear mixture of a small number of basis kernels, which limits expressivity and complicates optimization as the mixture size grows. In this work, we revisit dynamic convolution from a functional perspective and propose the Meta Convolution Layer (MCL), which directly models the convolutional kernel as an input-conditioned function W(x) realized via a high-order
polynomial expansion. Leveraging nested residual blocks inspired by deep
polynomial networks, MCL implements a structured
polynomial meta-network that generates a single input-adaptive kernel, thereby decoupling representational power from the explicit number of mixture kernels and alleviating training instability. MCL is a plug-in addition with standard convolutions and can be seamlessly integrated into both CNN and
transformer backbones. Experimental evaluation shows that adding MCL improves the Top-1 accuracy of Resnet- 18, Resnet-50 and ResNet-101 by 6.61%, 3.42% and 3.05% on the ImageNet
dataset. Moreover, the proposed method significantly boosts the accuracy of Resnet and Wide-Resnet variants on CIFAR-10 and CIFAR-100
datasets. Additionally, the proposed method outperforms previous methods on fine-grained visual classification tasks using Swin and ViT backbones. These results demonstrate that high-order
polynomial kernel generation is a powerful and scalable alternative to linear mixture based dynamic convolution.
Read the full abstract ▾
Statistics Applications
BayesJudge: Uncertainty-Aware Bayesian Meta-Evaluation of Human and LLM Judgments
Jiahao Zhang, Pengbin Feng, Chunlei Meng et al. (+2 more)
AI evaluation pipelines often produce conflicting judgments rather than clean labels. In pairwise LLM evaluation, this conflict is especially visible: disagreement can arise from ambiguous items, underspecified rubrics, heterogeneous or unstable human raters, or an LLM judge whose verdict changes when the response order is swapped. We propose BayesJudge, an online
Bayesian meta-evaluation layer for conflicting human-LLM judgment streams. For each comparison, BayesJudge estimates a panel-relative posterior verdict distribution over the two responses, with a tie or ambiguity state when such labels are available. At the same time, it estimates rater-specific human confusion
matrices and LLM presentation-order bias. The method uses tie-open labels to keep ambiguity observable and paired order-swapped judge calls to separate response quality from presentation effects. We formulate the exact online posterior recursion and use a scalable Rao-Blackwellized assumed-density SMC approximation for streaming inference. Controlled synthetic experiments demonstrate recovery of prespecified evaluator parameters and illustrate two protocol-level identifiability mechanisms: tie-open labels expose ambiguity mass, and order-swapped paired judgments separate item preference from position bias. On a real-world SummEval
dataset, BayesJudge successfully detects systematic presentation-order effects in LLM judge outputs, infers distinct expert and crowdworker behavior signatures without rater metadata, and produces posterior uncertainty estimates that correlate with human disagreement. Our code is available at https://anonymous.4open.science/r/BayesJudge-0879.
Read the full abstract ▾
All authors Jiahao Zhang, Pengbin Feng, Chunlei Meng, Hang He, Tianyong Hao
Machine Learning
Dynamics as Code: On Model Compression via Dynamic System
Fan Gao, Wei Su, Juntong Fan et al. (+4 more)
The escalating size of pretrained neural networks has rendered model compression a prerequisite for deployment under stringent
memory and compute constraints. With the irrational winding as an example, earlier work introduced a dynamic system (DS) paradigm that reconceptualizes compression as compact weight representation: high-dimensional parameters are encoded by the index of a trajectory produced by a dynamic system, from which the vector is recovered during decompression. This mechanism is fundamentally distinct from
pruning, quantization, knowledge
distillation, and low-rank decomposition. Along this direction, we prove that under a Diophantine condition, a finite trajectory of \(M = O(\epsilon^{-(d+\nu)})\) states in the irrational winding constitutes an \(\epsilon\)-net over the \(d\)-dimensional weight space, thereby linking state resolution, decompression error, and compression ratio in a predictable manner. Furthermore, we propose a generalized DS-based model compression framework by unifying four DS families—space-filling curves (
Hilbert, Peano, Morton/Z-order, Snake), chaotic systems (Lorenz), congruential and pseudo-random generators (LCG, PCG), and low-discrepancy sequences (Halton). Also, we introduce the KD-tree and coordinate-template acceleration to scale to large models as well as outlier identification to control the error. Experiments on ResNet-18 and Qwen2.5-1.5B/Qwen1.5-7B validate that DS-based compression achieves competitive compression ratios without post-hoc retraining, with controllable decompression error and flexible state-space design, establishing it as a principled and practical compression approach.
Read the full abstract ▾
All authors Fan Gao, Wei Su, Juntong Fan, Renfeng Peng, Hongyu Liu, Jinqiao Duan, Feng-Lei Fan
Distributed Computing
IAPRepair: In-Network Aggregation Enhanced Proactive Repair for Erasure-Coded Storage System
Lei Liu, Yong Wang, Junqi Chen et al. (+3 more)
Erasure-coded storage provides fault tolerance with substantially lower storage overhead than full replication, but repairing lost or at-risk blocks requires intensive cross-node data transfer. Existing reactive repair schemes start only after a failure, while proactive schemes can move data before failure but commonly treat migration, reconstruction, and network aggregation as loosely coupled operations. As a result, receiver?side bottlenecks, heterogeneous available bandwidth, and limited programmable-switch state continue to constrain repair paral?lelism. This paper presents IAPRepair, an in-network aggre?gation enhanced proactive repair framework for erasure-coded storage. IAPRepair jointly constructs each repair batch, assigns reconstruction providers and replacement nodes according to normalized transmission loads, schedules migration around the remaining receive capacity, and selectively enables in-network aggregation for reconstruction blocks that would otherwise over?load healthy nodes. The selective design reduces receiver-side traffic while retaining migration parallelism and respecting a configurable switch-resource budget. We implement IAPRepair with a Tofino programmable switch and 16 storage nodes, and evaluate it using both a prototype testbed and large-scale simu?lations. Across coding parameters, block sizes, node populations, and multiple STF-node scenarios, IAPRepair reduces repair time by at least 47.83% compared with the evaluated state-of-the-art methods. The results demonstrate that coordinating proactive repair decisions with in-network processing is an effective way to improve repair efficiency under bandwidth heterogeneity.
Read the full abstract ▾
All authors Lei Liu, Yong Wang, Junqi Chen, Yangfan Liang, Qian He, Baokang Zhao
Computer Vision
DiscoVL: Unveiling Disentangled C ross-Modal Representation Learning via Orthogonal Adversarial Regularization for V ision-Language Models
Mengping Dong, Jinbao Li and Fei Li
Pre-trained
vision-
language models excel across varied perception tasks, but adapting them to novel downstream settings without sacrificing generalization remains non-trivial. Existing parameter-efficient
prompt learning method often yields inconsistent representations and fails to account for semantic distribution shifts. In this work, we present DiscoVL, a disentangled cross-modal representation learning framework that couples orthogonal
adversarial regularization with structured cross-modal
alignment for
vision-
language models. To address the insufficient cross-modal interaction, our DiscoVL designs a multi-branch low-rank residual aligner that decomposes representations into subspaces and enables bidirectional cross-modal feedback between visual and textual streams at each layer. Furthermore, while conventional triplet constraints overfit features to class centroids, we design an orthogonal regularization for
adversarial triplet loss, which prevents centroid collapse and substantially boosts generalization. Evaluations on 15
benchmarks demonstrate that DiscoVL delivers consistent improvements over state-of-the-art methods for base-to-novel generalization, cross-
dataset evaluation, and few-shot learning
Read the full abstract ▾
Cryptography and Security
False Claims, Credible Images: A Red-Teaming Benchmark for Commercial Image Generators
Zeyu Ye, Yanchun Li, Sibei He et al. (+8 more)
Image-generation models can now produce text-rich, natural-looking visual artifacts that are hard to distinguish from real-world evidence, such as news reports and textbook pages. Yet, the same capability introduces a new risk: these models can just as easily fabricate visual misinformation. Even commercial models (e.g., GPT-Image-2) readily produce it. Curiously, we find that these models can recognize a claim as false when asked, yet still render that very claim as credible visual evidence. This discrepancy points to a blind spot in current
alignment: safeguards judge what an image shows, not what it asserts; however, existing red-teaming
benchmarks target conventional harmful content, such as violent or explicit imagery, and say little about where the
alignment boundaries lie for visual misinformation, especially in commercial models. To fill this gap, we introduce EpiReal-Bench, the first systematic
benchmark for evaluating visual misinformation risks in commercial image generators, comprising 10k false-claim
prompts and 10k corresponding generated images that span 10 real-world claim categories and 10 credible visual formats. We further introduce EpiReal-Attack, a skill-guided black-box optimization framework that uses Pareto-based selection and
multimodal feedback to identify commands that bypass
alignment safeguards while preserving visual realism, textual legibility, and semantic fidelity. Experiments on four commercial models reveal that more than 70% of false-claim
prompts elicit images that faithfully depict the corresponding misinformation, and EpiReal-Attack pushes this rate to 95%. Most worryingly, these models are only a click away, and their outputs are cheap to spread yet hard to disbelieve, leaving this dimension of
alignment largely unguarded.
Read the full abstract ▾
All authors Zeyu Ye, Yanchun Li, Sibei He, Meng Xie, Hangtao Zhang, Xianlong Wang, Li Zeng, Jiahao Chen, Yichen Wang, Junhui Wang, Ziqi Zhou
Computation and Language
Lapras: Latent Reasoning for Time Series Language Models
Yuliang Chen, Yu Yvonne Wu, Patrick Langer et al. (+11 more)
Time Series
Language Models (TSLMs) offer a promising path toward time series understanding by
reasoning over temporal signals and producing natural
language answers and explanations. A common approach is Chain-of-Thought (CoT), which generates step-by-step rationales linking relevant signal patterns to final answers. Although these models learn from reference CoT traces during post-training, generating faithful descriptions of input time series at inference remains challenging. Expressing high-dimensional, continuous temporal representations in discrete
language tokens may cause the model to neglect task-relevant patterns or describe them inaccurately. Because later
reasoning steps build on these descriptions, early errors propagate, leading to incorrect answers with plausible explanations that are inconsistent with the input signal. We propose Lapras (Latent Post-trained
Reasoning Across Series), a post-training framework that equips TSLMs with latent
reasoning. A model trained with Lapras reasons through a sequence of continuous thoughts in the joint time series-
language space, producing text only for the final answer. It learns this through teacher-student self-
distillation, where a teacher trained on CoT reference traces reasons explicitly through text. The student aligns its hidden states with the teacher's at the answer stage, transferring the teacher's
reasoning ability into its latent computation. We evaluate Lapras across four TSLM backbones on five time series question answering
benchmarks. Lapras improves average F1 by up to 10.79% over explicit CoT while generating 23.9x fewer
tokens. Lapras's continuous thoughts can also be decoded into readable
reasoning traces via standard
language decoding, preserving textual explanations. Together, these results highlight Lapras as a promising post-training paradigm for efficient, effective, and interpretable TSLM
reasoning.
Read the full abstract ▾
All authors Yuliang Chen, Yu Yvonne Wu, Patrick Langer, Arvind Pillai, Sudarshan Regmi, Martin Maritsch, Juncheng Liu, Robert Jakob, Thomas Kaar, Tess Z. Griffin, Lisa Marsch, Michael V. Heinz, Nicholas C. Jacobson, Andrew Campbell
Algebraic Topology
Braidings of Self-Equivalences and Bordism
Ian Hambleton and Kursat Sozer
Let $M$ be a closed, smooth or topological $n$-
manifold, with $n \geq 4$. We construct a homotopy highly cartesian square relating the space ${\mathcal E}(M (\ell))$ of homotopy self-equivalences of $M$ (in a suitable range) over the Postnikov $\ell$-sections of its stable normal microbundle, and an $(\infty + n)$-fold loop space representing an associated (normal) bordism theory. This implies the existence of braids of interlocking exact sequences involving the homotopy groups of ${\mathcal E}(M(\ell))$ and certain Lashof bordism groups, leading to a conceptual explanation and broad generalization of earlier work of Hambleton–Kreck for closed, oriented $4$-
manifolds.
Read the full abstract ▾
Algebraic Geometry
A Proof of Mongardi's Conjecture on Finite Symplectic Group Actions on Hyperkähler Manifolds of $K3^{[n]}$ Type
Jie Fu, Shihao Wang and Zhiwei Zheng
Let a finite group act faithfully by
symplectic automorphisms on a
manifold of $K3^{[n]}$ type for $n\geq2$. We prove Mongardi's
conjecture that the strict inequality $\mathrm{rk}(S)+\ell(A_S)<24$ holds for its associated Leech coinvariant
lattice $S$. The converse was already proved independently by Huybrechts and Mongardi: any Leech coinvariant
lattice $S$ satisfying the strict inequality and of rank at most $20$ is realized by such an action for some $n\geq2$. The key idea is to consider the minimum norms of vectors representing discriminant classes of $S$. The absence of numerical walls gives lower bounds for certain classes, while the structure of Leech
lattice implies upper bounds for all discriminant classes. Comparing these bounds reduces the proof to finitely many values of $n$. We exclude the remaining cases using further
lattice arguments, together with Höhn–Mason's classification of fixed-point sublattices of Leech
lattice, thereby proving Mongardi's
conjecture.
Read the full abstract ▾
Machine Learning
Cova-PINN: Cross-Domain Conservation Physics-Informed Neural Network for Fluid-Solid Conjugate Heat Transfer in Complex Geometries
Weizheng Zhang, Xunjie Xie, Hao Pan et al. (+1 more)
Multi-domain physics-informed neural networks (PINNs) flexibly model medium-specific representations to solve fluid–solid conjugate heat transfer (CHT). However, standard multi-domain PINNs enforce governing equations and interface conditions on separately sampled domain supports, which can yield plausible temperature fields but inaccurate end-to-end energy transfer and outlet temperatures. We propose Cova-PINN, a multi-domain PINN framework that aligns conservation support with thermal interaction paths in complex geometries. Cova-PINN jointly optimizes cross-domain composite control-volume balances at the local scale and paired-wall closure at the global exchanger scale. We evaluate Cova-PINN on four triply periodic minimal surface (TPMS) heat exchangers and a geometrically distinct DualMS design against CHT-specific, optimization-oriented, and complex-geometry PINN baselines under a common protocol. Relative to the closest baseline, MUSA-PINN-CHT, Cova-PINN reduces average outlet-temperature and device-level closure errors across the four TPMS
topologies by $37.7\%$ and $60.2\%$, respectively, while also improving full-field and heat-duty accuracy, with consistent gains on DualMS.
Read the full abstract ▾
All authors Weizheng Zhang, Xunjie Xie, Hao Pan, Lin Lu
Cryptography and Security
Accelerating HQC for Post-Quantum TLS 1.3 on x86 IoT Gateways
Jihoon Jang, Hyunju Park, Jebin Kim et al. (+2 more)
Post-
quantum TLS at an IoT gateway must protect many device connections without large increases in handshake delay, CPU cost, or network traffic. HQC provides code-based diversity beyond ML-KEM, but its computation and ciphertext sizes can increase these costs. We optimize HQC for x86 processors with AVX2, AVX-512, and the Galois Field New Instructions (GFNI), and integrate the resulting implementations into TLS 1.3. We extend branch-free Toom-Cook/Karatsuba multiplication to AVX-512, accelerate Reed-Solomon decoding with GFNI, improve Reed-Muller decoding, accelerate SHA3-512 and fixed-weight sampling, and port Frobenius additive FFT (FAFFT) multiplication to AVX-512 + GFNI for HQC-5. On an Intel Core i5-1135G7 processor, our AVX2 implementation reduces decapsulation by 11.1-21.5% over the fastest prior AVX2 results. Our AVX-512 implementation reduces key generation by 24.9-29.5%, encapsulation by 10.4-11.1%, and decapsulation by 20.6-26.1% relative to Cabral et al. across the three HQC parameter sets. We evaluate the effects of these implementations on TLS 1.3 handshake latency and server and client CPU costs, as well as the effects of round-trip time (RTT) and bandwidth. In local loopback TLS measurements, our AVX-512 HQC-5 implementation reduces handshake latency from 3.22 ms to 2.92 ms compared with Cabral et al. On a constrained path with 1 Mbit/s bandwidth and an added RTT of 50 ms, the HQC-5 handshake takes 243 ms, compared with 87 ms for ML-KEM-1024. This indicates that HQC public-key and ciphertext sizes, rather than implementation speed, determine the remaining handshake cost on constrained links.
Read the full abstract ▾
All authors Jihoon Jang, Hyunju Park, Jebin Kim, Seokhie Hong, Suhri Kim