Computation and Language
What Breaks Under Pruning in Smart Homes, and When? Evaluating LLM Degradation Across Architectures and Task Complexity
Congjing Zhang, Vashishtha Patil, Henning Lange et al. (+1 more)
Pruning can reduce the deployment cost of large
language models (LLMs), but its impact on context-grounded tool calling remains poorly understood. We systematically study
pruning-induced degradation in smart-home tool calling across four LLMs spanning dense
Transformer, dense hybrid, and mixture-of-experts (MoE) architectures, together with depth, width, hybrid, and expert
pruning methods. After post-
pruning supervised fine-tuning (SFT), we evaluate more than 19,500 instances from three smart-home
datasets. Beyond aggregate task accuracy, we characterize degradation along two dimensions: action components (i.e., operation, device, argument, and value) and task complexity. Our results show that dense models have narrow safe
pruning regions followed by sharp degradation, while MoE models tolerate substantially more
pruning.
Pruning degrades grounded specificity before schema-level intent, and aggressive dense
pruning can induce systematic over-refusal. These findings highlight the importance of evaluating
pruning beyond aggregate accuracy when selecting pruned LLMs for reliable tool execution.
Read the full abstract ▾
All authors Congjing Zhang, Vashishtha Patil, Henning Lange, Usman Aleem
Sound
LACE: Layer-Wise Compression for Dynamic Frame Rate Codecs
Thanapat Trachu, Samuele Cornell, William Chen et al. (+1 more)
Neural audio codecs are a key component in
speech language modeling. However, their high frame rates lead to long sequence lengths, increasing computational costs. Dynamic frame rate codecs mitigate this by reducing the effective frame rate using a compression step to merge multiple frames together. However, most prior methods either operate on single-codebook codecs or apply a single compression step before multi-layer quantization. This forces all quantization layers to share the same
segmentation boundaries, despite the residual
embeddings at different quantization layers exhibiting different rates of change over time. We propose LACE (Layer-Adaptive Codec Encoding), a dynamic frame rate codec that applies an independent compression step at each quantization layer, enabling layer-specific
segmentation boundaries. To use LACE
tokens in downstream text-to-
speech (TTS), we further introduce union
alignment and boundary anchor mechanisms to make durations consistent across layers while preserving compression benefits. Experiments on LibriTTS show that LACE offers a better rate-quality tradeoff than prior dynamic frame rate methods on the reconstruction task and improves TTS inference efficiency while maintaining competitive synthesis quality. Our code is released as part of the ESPnet3 codec recipe.
Read the full abstract ▾
All authors Thanapat Trachu, Samuele Cornell, William Chen, Shinji Watanabe
Machine Learning
ENCP: Episode-Normalized Conformal Prediction for Vision-and-Language Navigation
Vicky Feliren, A. Taufiq Asyhari and Muhamad Risqi U. Saputra
Uncertainty estimation for
Vision-
Language-
Navigation (VLN) models is a critical task since it can help identify ambiguous and unreliable predictions, enabling
agents to make safer
navigation decisions. As one of the most advanced uncertainty estimation frameworks, conformal prediction (CP) offers a promising approach for uncertainty estimation in VLN. However, given that VLN
agent requires a sequence of steps, standard calibration in conformal prediction fails to provide coverage guarantee it promises over a dependent, variable-length VLN episode. To this end, we propose Episode-Normalized Conformal Prediction (ENCP), which rescales a nonconformity score by the policy's residual confidence and calibrates one maximum score per episode. Under exchangeable calibration and test episodes, this construction covers the ground truth at every step with probability at least $1 - \alpha$, while allowing dependence among steps within an episode. Across four VLN policies and three nonconformity scores on R2R and REVERIE
dataset, ENCP meets all reported empirical step-coverage targets on the seen-to-unseen evaluation. These results demonstrate that ENCP can provide model-agnostic uncertainty estimates, which might be useful for determining when a VLN
agent should defer to a more capable predictor, including human assistance.
Read the full abstract ▾
Artificial Intelligence
Verifiable Social Reasoning for LLM Assistants
Amir Taubenfeld, Zorik Gekhman, Avigail Grinstein-Dabush et al. (+6 more)
LLM assistants are widely used for daily social advice, yet evaluating their social
reasoning in such consultation settings remains challenging since (i) it requires setups where the assistant learns about social situations from subjective user narratives, and (ii) social properties, such as others' intentions, typically lack verifiable ground truth. To address these challenges, we introduce Fuse, a multi-
agent simulation framework for studying user-mediated social
reasoning. In Fuse, a target
agent with a hidden motive interacts with other
agents including one representing the user, who then consults the evaluated assistant to infer the target's motive, providing verifiable ground truth by construction. Simulation faithfulness is validated through a human study with 24k annotations. We apply Fuse to 12 LLMs and demonstrate its analytical utility by systematically isolating key factors, showing that (i) user mediation compounds the inherent difficulty of social
reasoning; (ii) LLMs exhibit systematic sensitivity to biased user framing; (iii) models can require more details than humans need to reach a correct prediction; and (iv) longer conversations do not always improve performance despite providing opportunities for clarifying questions. We open-source Fuse and a
dataset with 21k examples.
Read the full abstract ▾
All authors Amir Taubenfeld, Zorik Gekhman, Avigail Grinstein-Dabush, Itay Laish, Ariel Goldstein, Marian Croak, Avinatan Hassidim, Yossi Matias, Amir Feder
Numerical Analysis
On the Existence of Pressure-Equilibrium-Preserving Numerical Fluxes for Supercritical Fluids
Robin Ben Klein
In this work we propose a new existence
theorem for numerical-flux functions for supercritical fluids that are pressure-
equilibrium-preserving (PEP). In particular, we characterize the existence of consistent numerical-flux functions that satisfy an algebraic PEP property. Our theory links the existence of PEP schemes to
geometric properties of the equation of state describing the thermodynamics of the fluid. When these
geometric properties fail for a pair of states on the same isobar, no PEP schemes of the considered form can exist on a domain containing those states. In our analysis the equation of state itself can be fully general only needing to satisfy some fundamental thermodynamic principles. Recently, PEP compatibility conditions for general equations of state have been derived [1] that rely on the existence of certain thermodynamic derivatives which are not guaranteed to be defined under fundamental thermodynamic principles. The
geometric conditions in our theory do not depend on the existence of these derivatives and recover the recent compatibility conditions in the case that these derivatives are defined. Finally, using numerical experiments we demonstrate for two supercritical fluids that our existence conditions are restrictive and thus that no PEP schemes of the form we consider exist on the domain we specify for these fluids. Using our
geometric perspective, we also shed light on mechanisms by which PEP schemes can develop numerical issues, which we also demonstrate using numerical experiments.
Read the full abstract ▾
Machine Learning
FreqSpaNet: Frequency and Spatial Learning of SFPF for Physical Layer Hardware Integrity Detection
Xiaoxuan Huang, Jinlong Xu, YiZhe Wang et al. (+3 more)
Unauthorized hardware replacement can preserve a wireless device's logical identity while altering its physical implementation, posing a challenge to hardware integrity verification. Spatio-frequency polarization fingerprints (SFPFs) capture device-dependent responses across multiple frequencies and directions, but their frequency and spatial dimensions exhibit different structural dependencies. We propose FreqSpaNet, an SFPF representation learning network for open set hardware anomaly detection. A frequency branch captures local variations among neighboring frequencies, while a geometry-aware spatial branch models directional relationships using angular information. The two representations are combined through adaptive fusion, and complementary pretraining further captures shared information while preserving the distinct characteristics of the frequency and spatial representations. Experiments show that FreqSpaNet achieves a mean AUROC of 96.31%, 9.05 points above the baseline. Results under seven hardware replacement scenarios further verify the effectiveness of FreqSpaNet.
Read the full abstract ▾
All authors Xiaoxuan Huang, Jinlong Xu, YiZhe Wang, Meng Zhang, Xian Li, Yuying Bian
Artificial Intelligence
LimiX-2: A Contextual Mechanism Network Towards General Structured-Data Intelligence
Xingxuan Zhang, Gang Ren, Hao Yuan et al. (+57 more)
We introduce LimiX-2, a new model in the LimiX family, developed through model and data scaling guided by our previously established scaling laws. LimiX-2 adopts the Contextual Mechanism Networks (CMNs) paradigm and is pretrained with Context-Conditional Masked Modeling (CCMM). CMNs shifts the organizing principle of in-context learning from target-centric prediction to mechanism-oriented joint modeling. Rather than centering the network on the $p(y \mid x, D_{\mathrm{context}})$ objective of conventional tabular PFNs, it is designed around learning $p(x, y \mid D_{\mathrm{context}})$, a context-dependent representation of the joint structure underlying data generation. Pretraining uses synthetic
datasets generated by structural causal models (SCMs) spanning diverse graph structures, functional mechanisms, and observation processes. Evaluations on TabArena, TALENT, and BCCO show that LimiX-2 outperforms current
dataset-specific models and tabular foundation models. Beyond predictive performance, the CMN paradigm also promotes causal awareness in LimiX-2: its feature
attention encodes direct causal relationships, enabling accurate causal skeleton recovery.
Read the full abstract ▾
All authors Xingxuan Zhang, Gang Ren, Hao Yuan, Hao Zou, Hongze Tan, Hui Wang, Jianhao Song, Jiansheng Li, Jiayao Zhang, Jinghan Zhang, Kaifang Li, Lang Mo, Li Mao, Mingchao Hao, Nuo Xu, Rui Ding, Ruiji Zhang, Shuyang Li, Siyu Mei, Tianyang Zhang, Weiyang Mu, Yancheng Dong, Yongxian Wei, Yuan Xue, Yuanrui Wang, Yue He, Zijia Yang, Ziyun Li, Dongzhe Li, Fuqiang Wang, Jiandong Liu, Jiawei Chen, Jiaxin Du, Kaijie Cheng, Kehan Li, Lei Sun, Linjun Zhou, Ningbo Dai, Qi Wang, Renzhe Xu, Shaoxing Du, Shumeng Yang, Wang Lu, Wenjing Chu, Xiannan Huang, Xiaoyu Lin, Xing Ai, Xinyan Han, Xuanyue Li, Xuanyue Su, Xukun Zhang, Yan Lu, Yaxin Zhang, Yi Qin, Yifei Huang, Yihan Xu, Yongle Lv, Yuanyuan Jiang, Yushan Han, Peng Cui
Data Structures and Algorithms
Stuffed IBLTs: Optimal Linear Multiset Sketches
Jonas Klausen, Rasmus Pagh and Stefan Walzer
A linear sketch is a randomized linear mapping of a vector $v$ to a lower dimensional sketch vector, designed to preserve relevant information about $v$. We consider sketches of vectors $v \in Z^u$ (for $u \in N$), designed for exact recovery of $v$ from its sketch. Concretely, our Stuffed IBLT is a linear sketch configured with a capacity $n \in N$ and a multiplicity limit $L \in N$ and will recover $v$ with high probability whenever $||v||_0 \leq n$ and $||v||_\infty \leq L$. The sketch can be maintained efficiently under unrestricted updates to $v$, i.e., $v$ is not subject to any constraints in between decoding requests. This makes the sketch useful for streaming
algorithms and for solving the (multi)set reconciliation problem. For any positive constants $c$, $\epsilon$, and for large enough $n$ and $u \geq n^{1+\Omega(1)}$, the space usage of a Stuffed IBLT is within a factor $1+\epsilon$ from the information-theoretic optimum while allowing updates in constant time, and decoding in time $O(n)$ with failure probability $n^{-c}$. This improves the space/time/error probability trade-off over all prior constructions with similar functionality, including the Invertible Bloom Lookup Table (IBLT). The performance of the Stuffed IBLT is essentially the best we could hope for, up to the dependence on $c$ and $\epsilon$. We make the dependence on these parameters explicit, and further show a lower bound demonstrating that the dependence on $c$ is optimal within the class of peeling-based approaches. Our improvement comes from a careful combination of Walzer's spatial coupling technique (SODA '21), the purity heuristic of Houen, Pagh, and Walzer (SOSA '23), and backyarding (Belazzougui, Kucherov, and Walzer, ESA '24; Fleischhacker, Green Larsen, Obremski, and Simkin, ICALP '24), allowing us to eliminate bottlenecks of past approaches.
Read the full abstract ▾
Human-Computer Interaction
Quick-View Takeaways: How Does Title Framing Influences Pattern Identification in Line Charts?
Jasmine Lim, Tapendra Pandey, Arran Zeyu Wang et al. (+1 more)
Visual data communication in digital media is increasingly characterized by short
attention spans and snapshot-based viewing, often employing line charts to convey trends and patterns. Among all visual elements, titles are crucial ones that can shape how viewers interpret visual information and form chart takeaways. In this study, we examine how title characteristics, particularly title word count and intended message, influence people's pattern identification in single-class line charts. Participants viewed 50 line charts collected from online news media and identified the pattern they perceived. Our results demonstrate that both title word count and intended message significantly influence viewers' pattern identification. Our findings highlight the importance of title design in shaping chart takeaways and effective visualization communication.
Read the full abstract ▾
All authors Jasmine Lim, Tapendra Pandey, Arran Zeyu Wang, Ghulam Jilani Quadri
Robotics
Dissecting Motion-Prior Regularization for Data-Scarce Robotic Insertion
Ning Hu, Shuai Li and Jindong Tan
This study asks whether training-time motion-prior regularization can improve insertion success when a
diffusion policy is learned from only 15 demonstrations. Minimum jerk discourages abrupt changes in predicted translational acceleration; speed-
curvature regularization instead couples movement speed to path geometry. These are candidate mechanisms for task completion, not safety guarantees. We compare the priors individually and jointly, neither prior, and generic smoothness, with 80 real-
robot trials per setting pooled over four recorded condition classes. Joint and minimum-jerk-only settings each achieved 70/80 successes (87.5%), versus 69/80 (86.3%) for speed-
curvature only, 66/80 (82.5%) for neither prior, and 67/80 (83.8%) for generic smoothness. Success rates and Wilson 95% confidence intervals are visualized for direct comparison. Joint regularization exceeded neither by 5.0 percentage points but provided no observed gain over minimum jerk alone. The results motivate minimum jerk as the simpler candidate for replication, without establishing synergy, biomechanical specificity, improved safety, or distribution-shift robustness.
Read the full abstract ▾
Machine Learning · Cross-listed from math.OC
Bridging the Gap Between Homogeneous and Heterogeneous Asynchronous Optimization Is Surprisingly Difficult
Alexander Tyurin
Modern large-scale machine learning tasks often require multiple workers, devices, CPUs, or GPUs to compute
stochastic gradients in parallel and asynchronously to train model weights. Theoretical results typically distinguish between two settings: (i) the homogeneous setting, where all workers have access to the same data distribution, and (ii) the heterogeneous setting, where each worker operates on different data distributions. Known optimal time complexities in these settings reveal a significant gap, with far more pessimistic guarantees in the heterogeneous case. In this work, we investigate whether these pessimistic optimal time complexities can be overcome under different assumptions. Surprisingly, we show that improvement is provably impossible under widely used first- and second-order similarity assumptions for any randomized
algorithm. We then turn to the interpolation regime and demonstrate that the weak interpolation assumption alone is also insufficient. Finally, we introduce a minimal combination of irreducible assumptions, strong interpolation and the local Polyak-Lojasiewicz condition, to derive a new time complexity bound that matches the dependence on worker computation times in the best-known result in the homogeneous setting, without requiring identical data distributions.
Read the full abstract ▾
Computer Vision
Det-LIME: Detector-Aware, Multi-Instance Local Interpretable Model-Agnostic Explanations for Automated Marine Mammal Detection
Jiayi Zhou, David W. Johnston and Brinnae Bent
Despite the rapid uptake of black-box object detectors in marine mammal research and monitoring, explainability techniques are rarely integrated into conservation workflows. Furthermore, most classification-oriented explainability tools are ill-suited to detection tasks involving imagery of social organisms or those with colonial life histories, as they ignore multiple detections within a scene and produce single-instance outputs that blur evidence across individuals. These methods also generate low-resolution, often biologically irrelevant visuals, limiting their utility for debugging, targeted data augmentation, and refined data collection. We proposed Det-LIME, a detector-aware, multi-instance adaptation of Local Interpretable Model-Agnostic Explanations (LIME) that produced instance-specific, box-aligned explanations by combining per-detection weighting, a proximity kernel that emphasizes regions near each box, and Intersection-over-Union-based matching to track the same instance across perturbations. We evaluated Det-LIME on aerial drone imagery for harbor seal detection, with an additional seabird case study to assess generality, and compared it with vanilla LIME, Stabilized LIME, Deterministic LIME, and gradient-based attribution methods. Using the Attribution Ratio and Max Saliency Hit Rate metrics, we showed that Det-LIME consistently improved multi-instance attribution. In practice, these higher-resolution, instance-aware explanations provide insight into model outputs and support post-processing, debugging, and actionable improvements in modeling and data collection or augmentation.
Read the full abstract ▾
Machine Learning · Cross-listed from cond-mat.dis-nn
Bias-Induced Crossover in Absolute Capacity of Dense Associative Memory
Yuto Sakurai, Takeaki Shimokawa, Kazunori Iwata et al. (+1 more)
The absolute capacity of dense associative
memory has mainly been analyzed for unbiased patterns. Here we examine the effect of bias in centered binary patterns under the Krotov-Hopfield single-site criterion $P_{\mathrm{error}}=1/N$, where $P_{\mathrm{error}}$ is the probability that a single-site flip lowers the energy of a stored pattern and $N$ is the number of
neurons. Each pattern component takes $1-q$ with probability $q$ and $-q$ otherwise, where $0<q\le1/2$. For
polynomial interactions of order $n$, a signal-to-noise analysis gives an absolute capacity of order $N^{n-1}/\ln N$ at $q=1/2$. For fixed $q<1/2$, however, the capacity is $O(N^{n/2})$ for even $n\ge4$ and $O(N^{(n+1)/2})$ for odd $n\ge5$. For $n=3$, both the unbiased and fixed-bias capacities remain $O(N^2/\ln N)$. For $n\ge4$, these different asymptotic forms imply a nonuniform large-$N$ limit near $q=1/2$. Asymptotic matching predicts a bias-induced crossover in the region $1-2q=O(\ln N/N^{\lfloor n/2\rfloor-1})$. The crossover originates from a bias-dependent crosstalk mean that reduces the stability of sites carrying the more frequent value $-q$. Computer simulations are compared with the finite-size conditioned-Gaussian predictions. An activity-dependent control potential that cancels the conditional crosstalk mean restores the $N^{n-1}/\ln N$ capacity for fixed $0<q<1/2$ within the conditioned-Gaussian approximation.
Read the full abstract ▾
All authors Yuto Sakurai, Takeaki Shimokawa, Kazunori Iwata, Kazushi Mimura
Artificial Intelligence
JustFit: Just-in-Time State Management for Local LLM Serving
Yuhua Chen
Local
agents need
memory for model execution and working history. We present JustFit, an MLX runtime that coordinates their overlapping allocations: KVExec executes and checkpoints four-bit KV with bounded workspace, PhaseSwap loads phase-dependent components, and StateTrans preserves history across execution modes. With Qwen3.8-27B MXFP4 on a 24 GiB M4 Pro MacBook, JustFit reaches 327,680 retained positions across two requests, 10.67 times the evaluated baseline's 30,720-position single-request record. One request completes the full 262,144-position native window at median 5.986
tokens/s. Each shape completes 16,384 outputs per request in three fresh processes: B1 uses one cold build and two prefix extensions; B2 uses three ordered prefix extensions (Section 4). Image encoding can proceed while preserving a live 196,608-input text request. A controlled, repetitive 32K+6K workload reaches median 18.284
tokens/s at 15,626 MiB; a separate AIME 2026 evaluation scores 29/30. Coordinating execution and state lifetimes makes longer histories feasible on personal hardware.
Read the full abstract ▾
Machine Learning
Iterative Calibration and Learning: Mitigating Teacher Bias in LLM Distillation without Target-Domain Reward Feedback
Haichen Hu, Yuheng Zhang and David Simchi-Levi
Large
language model (LLM)
distillation can propagate teacher bias under
covariate shift when target reward feedback is unavailable. We propose Iterative Calibration and Learning (IterCaL), which alternates source-reward teacher calibration through
token-level branching with student learning on target
prompts. We prove
polynomial convergence in expected average KL to the oracle student, the reference-regularized target-reward maximizer within a potentially misspecified student class; a quantified source-to-target calibration discrepancy gives an explicit additive true-return term. Our separation shows that regularized direct teacher matching retains positive oracle-KL error at its population optimum for every shared reference penalty, even when the teacher is stronger than every student. On the same instance, IterCaL converges to the oracle using source rewards alone.
Read the full abstract ▾
Numerical Analysis · Cross-listed from math.PR
Convergence in Hölder norms for Markovian approximations of stochastic Volterra equations
Noé Corneille, Kristin Kirchner and Pietro Pezzoli Frigerio
We bound the difference between two
stochastic Volterra processes with identical Lipschitz coefficients but different kernels. For non-convolution kernels, we establish estimates in $C^0([0,T];L^p(\Omega))$, $p\geq 2$, and for convolution kernels in $L^p(\Omega;L^q(0,T))$, $q \in [1,p]$, and $C^\beta([0,T];L^p(\Omega))$, $L^p(\Omega;C^\beta([0,T]))$, where the range of the Hölder exponent $\beta \in (0,1]$ is the maximal permitted by the regularity of the processes. For the fractional kernel, we then construct Markovian approximations whose error we show to decay as $e^{-a\sqrt{N}}$ in the aforementioned norms, using an $N$-node quadrature based on sinc methods. Numerical experiments for the fractional Brownian motion verify our findings.
Read the full abstract ▾
Multiagent Systems
Decomposition Buys Integrity, Not Yield
Rong He
Multi-
agent systems split a task across a tree of
agents and justify the split with folklore: smaller contexts, cleaner separation, parallelism. We ask what the split does to how much of what the leaves discover reaches the root. Model a decomposition as a tree in which an
agent handed $b$ items keeps any one with probability $r(b)$. If $r(b)=1/b$, every tree delivers exactly one finding, for every task size and every shape; we verify this to $2.4 \times 10^{-15}$ on 20,000 random irregular trees. If $r(b)=Cb^{-\delta}$, a depth-$k$ tree over $N$ findings yields $C^k N^{1-\delta}$: task size and architecture separate, and architecture contributes only $C \le 1$ per level, so flat is optimal for yield and no arrangement of
agents escapes the exponent $\delta$. On 600 production deep-research traces $\delta = 0.34$ [0.30, 0.38], by three identifications that do not share a failure mode. At a hop where item boundaries come from the tool rather than a text heuristic, and where $b=1$ occurs 550 times, $C = 0.571$ [0.527, 0.615] is observed rather than extrapolated, over 16,082 hops. A tier also costs
alignment: on 1,012 annotated multi-
agent traces one brief in sixteen goes off-target, giving $\mu = 0.939$ and a per-tier penalty $C\mu = 0.536$. Depth is bought on two other axes. The root context is the only state that persists and the only one that cannot cheaply forget, and depth cuts its exposure from $N$ items to $N^{1/k}$. Depth is also cheaper: production flat
agents bill as $N^{1.39}$, not the $N^2$ an append-only context predicts, and at equal spend two tiers overtake flat at 403 findings. Across every parameter we measured the model says 0.7% to 11.3% of production sessions are worth delegating, against 7.8% that do. A hazard model on 743,819 production tool calls finds that delegation does not respond to a filling context and is instead an opening move.
Read the full abstract ▾
Robotics
Gaussian Processes for Modelling Spatial Fields with Robot Swarms
Guillermo Legarda Herranz, Gianpiero Francesca and Mauro Birattari
Robot swarms, by virtue of their decentralised architecture, are a natural tool for scalable, robust modelling of spatial fields, such as water temperature, wind velocity, or terrain elevation. However, existing methods rely on external positioning systems that allow each
robot to determine its own position in space. Here, we introduce location-unaware Gaussian process regression (LU-GPR) as a solution to the modelling of spatial fields in the absence of such positioning systems. LU-GPR allows each
robot to infer the posterior mean and variance of the field in space, while simultaneously agreeing on a common frame of reference with its peers, using only local sensing and communication. We propose an online
algorithm that allows each
robot to consistently infer local estimates as its local frame of reference converges to the common one. By means of a product of experts model, each
robot also combines the estimates of its peers with its own to obtain a global model. Our results show that LU-GPR scales well with the number of
robots and is robust to limited communication ranges. We also demonstrate how it can be used in real-world monitoring scenarios to estimate the flow of an evacuating crowd.
Read the full abstract ▾
Computer Vision
Tables Decoded: DELTA for Structure, TARQA for Understanding
Jahanvi Rajput, Dhruv Kudale, Saikiran Kasturi et al. (+2 more)
Table understanding is a core task in document intelligence, encompassing two key subtasks: table reconstruction and table visual question answering (TabVQA). While recent approaches predominantly rely on
vision-
language models (VLMs) operating on table images, we propose a more scalable and effective alternative based on structured textual representations. These representations are easier to process, align more naturally with LLMs, and eliminate the need for
language-specific visual encoders, making them particularly suitable for
multilingual documents. We present DELTA, which separates physical structure recognition, logical structure recognition, and OCR to extract both layout and content accurately. DELTA outputs tables in Optimised Table Structure
Language (OTSL), a compact and unified format that encodes
cell arrangements and textual content. On table structure recognition (TSR), DELTA achieves TEDS- Structure scores comparable with state-of-the-art methods across FinTabNet, PubTabNet, and PubTables-1M. We further establish its robustness on non-English tables through our curated Hindi
benchmark, TORQUE. Building on this, we introduce TARQA, an LLM fine-tuned on OTSL sequences. Our approach yields gains of 9.3 p.p. on WTQ (TabQA) and 9.2 p.p. on FinTabNetQA (TabVQA), respectively. On TORQUE, our method ranks second among all VLMs and DELTA + LLM variants. We release our code, models, and
benchmark at: https://github.com/Tihiitborg/Tables-Decoded
Read the full abstract ▾
All authors Jahanvi Rajput, Dhruv Kudale, Saikiran Kasturi, Utkarsh Verma, Ganesh Ramakrishnan
Human-Computer Interaction
How Does Title Framing Influence Pattern Identification in Line Charts?
Jasmine Lim, Tapendra Pandey, Arran Zeyu Wang et al. (+2 more)
Visual data communication in digital media is increasingly characterized by short
attention spans and snapshot-based viewing, often employing line charts to convey trends and patterns. Among all visual elements, titles are crucial elements that can shape how viewers interpret visual information and form chart takeaways. In this study, we examine how title characteristics, particularly title word count and intended message, influence people's pattern identification in single-class line charts. Participants viewed 50 line charts collected from online news media and identified the pattern they perceived. Our results demonstrate that both title word count and intended message significantly influence viewers' pattern identification. Our findings highlight the importance of title framing in shaping quick-view pattern takeaways and supporting effective visualization communication.
Read the full abstract ▾
All authors Jasmine Lim, Tapendra Pandey, Arran Zeyu Wang, Sungahn Ko, Ghulam Jilani Quadri
Computer Vision
ORCA: Occlusion-Aware Refinement and Completion for Novel View Synthesis
Weronika Jakubowska, Maciej Zięba and Przemysław Spurek
Novel-view synthesis from a single image is a fundamentally ambiguous problem. As the camera moves away from the input viewpoint, previously hidden regions become visible, exposing missing geometry and holes in the reconstructed scene. Existing methods often rely on generative models to complete such regions. However, many of these artifacts are small gaps near depth boundaries and do not require generating new scene content. In order to eliminate expensive process of generating image we introduce ORCA, an occlusion-aware method for reconstructing and completing explorable 3D scenes from a single image. ORCA first introduces 3D structure into a Gaussian-anchor representation using
monocular depth while preserving the original camera-ray correspondence. During scene exploration, missing regions are handled based on their size and structure. Small disocclusions are repaired using RGB-D information already available in the reconstruction, while generative inpainting is reserved for larger regions that cannot be reliably recovered from the scene. New Gaussian anchors are added and optimized locally without modifying the existing representation. By reducing unnecessary reliance on generative inpainting, ORCA limits generation-induced
hallucinations and better preserves the content and structure of the original scene. On DIV2K, ORCA improves novel-view quality over VistaDream across all reported metrics, increasing MUSIQ from 61.60 to 68.71 and CLIP-IQA from 0.474 to 0.574. These results show that many novel-view artifacts can be repaired effectively by reusing information already present in the reconstructed scene.
Read the full abstract ▾
Social and Information Networks
Graphlets as structural fingerprints of complex networks
Anna Pidnebesna, David Hartman, Aneta Pokorna et al. (+2 more)
Complex networks are often compared using selected graph-theoretical measures that capture a selected set of properties with effects ranging from local to global, such as degree, clustering or betweenness centrality. Here we introduce a structural fingerprinting framework based on graphlets: small rooted subgraphs whose distributions provide a systematic description of local-to-mesoscale
topology. Across synthetic networks generated from several random graph models, graphlet fingerprints capture parameter-dependent structural differences, outperform standard graph-theoretical measures, and identify even subtle local patterns driving discrimination. We then apply the framework to empirical resting-state functional connectomes, documenting that while graphlets show superior sensitivity also to controlled topological perturbations of
brain connectivity, specifically in schizophrenia-control classification they perform only comparably to classical graph-theoretical features. This is in line with the notion that schizophrenia-related alterations are dominated by spatially localized connectivity changes rather than general topological reorganization. Altogether, the generative modeling, targeted perturbations and real-world neuroimaging classification challenge position graphlets as flexible structural fingerprints of complex networks, while carefully outlining their strength and weaknesses compared to more classical graph theoretical features.
Read the full abstract ▾
All authors Anna Pidnebesna, David Hartman, Aneta Pokorna, Daniel Trlifaj, Jaroslav Hlinka
Computer Vision
BrainFocus: EEG-Guided ROI Selection for Efficient Vision-Language Models
Yihui Peng, Guorui Lu and Qinyu Chen
Vision-
language models (VLMs) achieve strong visual question answering (VQA) performance, but processing large cluttered images is computationally expensive when only a small region is relevant. Electroencephalography (
EEG) signals, which capture human neural responses to visual stimuli, can provide a human-derived semantic cue about the region of interest (ROI). However,
EEG-guided visual category decoding remains imperfect, making direct ROI routing unreliable. In this work, we propose BrainFocus, a reliable
EEG-guided efficient VLM framework for VQA. An
EEG classifier predicts a target category, and a YOLO detector localizes the matching ROI. The VLM receives the cropped ROI only when both predictions pass confidence thresholds; otherwise, it processes the full image. For evaluation, we build on
EEG-ImageNet to construct a 40-class
benchmark comprising generated cluttered images and real object-centric images, with target-ROI annotations and 600 English visual question-answer pairs. Across Qwen3.5-VL 2B, 4B, and 9B models, BrainFocus improves VQA accuracy by 4.14-9.87 percentage points (pp) on cluttered scenes while reducing input
tokens and total
tokens by 23.2%-39.4% and 23.2%-39.3%, and end-to-end floating-point operations (FLOPs) by 23.2%-39.5%. These results demonstrate that
EEG can guide efficient VLM inference even when its semantic decoding is imperfect.
Read the full abstract ▾
Machine Learning
Reduced-Space Multi-Fidelity Bayesian Optimization of Process Simulation Models
Niki Triantafyllou, Andrea Bernardi and Maria M. Papathanasiou
Optimizing industrial process flowsheets is often computationally prohibitive due to the high cost of rigorous simulations and the curse of dimensionality inherent in complex design spaces. To address these challenges, we present a reduced-space multi-fidelity
Bayesian optimization (RS-MFBO) framework designed for high-dimensional, expensive black-box functions. The approach integrates Global Sensitivity Analysis (GSA) for dimensionality reduction with a fidelity-augmented Gaussian process that captures correlations between low-cost approximations and expensive high-fidelity evaluations. A cost-aware acquisition strategy, augmented with cooldown and promotion mechanisms, adaptively guides the allocation of samples across fidelities. The framework is validated on two distinct industrial process simulators: a plasmid DNA bioprocess in SuperPro Designer and a green fuel synthesis plant in Aspen HYSYS. Results across diverse economic and physical objectives demonstrate that the proposed method substantially reduces the number of high-fidelity simulator evaluations while maintaining competitive optimization performance compared to single-fidelity baselines. These results highlight RS-MFBO as a scalable, simulator-agnostic approach for cost-constrained black-box optimization.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from quant-ph
Evaluating Verified Autonomy in Quantum Engineering
Naixu Guo, Changhao Li, Siyu Cheng et al. (+6 more)
Reliable
quantum engineering is essential for turning
quantum phenomena into practical technologies. As
quantum platforms grow in scale and complexity, their characterization and operation require increasing human effort and coordination. Scientific artificial intelligence
agents, which can plan experiments, operate instruments, and analyze observations, offer a promising route towards autonomous
quantum engineering. Yet whether current
agents can perform reliably in this setting has not been systematically established. To fill this gap, we developed
Quantum-Harbor, a virtual laboratory that provides a controlled execution environment for
agents to interact with
quantum systems. This design enables direct verification of both the actions taken and the conclusions drawn. Building on this framework, we introduce QIQCBench, a
benchmark of $49$ expert-authored tasks spanning multiple layers including calibration and control, error correction and compilation, sensing and networking. Across $17$ frontier agentic systems, QIQCBench reveals wide variation in verified performance. These results expose a substantial gap between demonstrating capability and achieving reliable operation, and establish
Quantum-Harbor as a foundation for measuring progress towards verified autonomy in
quantum engineering.
Read the full abstract ▾
All authors Naixu Guo, Changhao Li, Siyu Cheng, Qicheng Tang, Binzhao Luo, Bikun Li, Yuxuan Du, Shihao Ru, Jiaqi Cai
Computation and Language
Right Tool, Right Job: Native-Language Evaluation, Tokenizer Sensitivity, and Methodological Findings from a French-Only BabyLM
Adam Zachary Wasserman and David Beauchemin
We submit MéTRON-FR, a 125M GPT-2 pretrained on 92.47M words of French, to the BabyLM 2026 Strict track. It scores 85.97 +/- 0.17% on QFrBLiMP (a native Quebec-French
benchmark of grammatical minimal pairs) and 62.80% on the BabyLM-weighted leaderboard. A cross-lingual GLUE (General
Language Understanding Evaluation) protocol that combines French task-data translation with rank-16 LoRA (Low-Rank Adaptation) produces a sharp task-type gradient: relational tasks gain measurably, while world-knowledge tasks regress. Bilingual Lexicon Induction aligns the French
embeddings to GPT-2 at p@1 = 68.84 +/- 8.61%, 18X above chance, suggesting cross-lingual
alignment tracks acquired grammatical competence rather than training duration. An ablation study shows that single-
token zero-shot scoring is dominated by tokenizer and template artifacts at the child scale, motivating tokenizer-swap sensitivity, placebo-controlled prompting, and native-
language minimal-pair
benchmarks as standard diagnostics.
Read the full abstract ▾
Human-Computer Interaction
CareMirror: Bringing Caregiver Wellbeing into the Dementia Care Ecosystem
Jiayue Melissa Shi, Ethan Nguyen, Drishti Goel et al. (+7 more)
Family caregivers of people living with dementia shoulder emotional and practical responsibilities, yet their own wellbeing often remains peripheral to dementia care. We built CareMirror, an envisioned caregiver wellbeing ecosystem with interconnected caregiver- and clinician-facing interfaces for longitudinal reflection, personalized support, and caregiver-controlled sharing with clinical care. We conducted semi-structured interviews with 14 caregivers, using CareMirror as a design probe to examine how they perceived this ecosystem and what expectations, concerns, and boundaries emerged around clinical connection. Caregivers valued
attention to their wellbeing, longitudinal awareness, context-sensitive support, and clinical visibility when it could lead to meaningful follow-up. However, repeated reflection could become burdensome or emotionally difficult, automatic clinical sharing could inhibit candid disclosure, and participants wanted control over what information entered clinical care. They also expected AI to support reflection and communication without replacing caregiver voice or clinician judgment. We contribute design considerations for proactive, clinically connected caregiver wellbeing support.
Read the full abstract ▾
All authors Jiayue Melissa Shi, Ethan Nguyen, Drishti Goel, Upasana Natarajan, Shashwat Srivatsa, Daniel S. Brown, Violeta J. Rodríguez, Dong Whi Yoo, Ravi Karkar, Koustuv Saha
Information Theory
Analytical Channel Modeling and Stability Aware Optimization of Optical Inter Satellite Links
Hossein Safi, Ziheng Wang, Stijn Mast et al. (+2 more)
Optical inter-satellite links (OISLs) are key enablers for high-capacity space networks and next-generation satellite constellations. However, their extreme directionality makes link reliability highly sensitive to platform-induced pointing jitter, which causes random misalignment between the transmitter and receiver beams. In this paper, we develop a tractable closed-form statistical channel model for point-to-point OISLs subject to independent pointing errors at both terminals. Accurate Gaussian main-lobe approximations are applied to the transmitter far-field pattern and receiver coupling efficiency. This transforms the diffraction-based channel response into closed-form expressions for the channel-gain distribution, outage probability, and ergodic capacity. The analytical results are validated through Monte Carlo simulations and used to study the impact of terminal stability, beam divergence, and link margin on OISL performance. The results show that outage probability is governed by the weaker terminal in terms of pointing stability, while improving only the stronger terminal provides minimal additional benefit. In contrast, the ergodic-capacity penalty depends on the combined stability of both terminals, revealing a fundamental distinction between reliability and throughput metrics. The proposed framework provides practical design guidelines for selecting beam parameters and specifying pointing and tracking requirements under varying levels of platform instability.
Read the full abstract ▾
All authors Hossein Safi, Ziheng Wang, Stijn Mast, Harald Haas, Iman Tavakkolnia
Robotics
Hamilton-Jacobi Reachability for Hybrid Systems: Unified Goal-Driven Control with Safety Guarantees
Javier Borquez, Shuang Peng and Somil Bansal
Hybrid dynamical systems provide a powerful modeling framework for robotic systems, particularly in contact-rich environments. However, ensuring safety and performance in such systems remains challenging due to the intricate coupling between continuous dynamics and discrete mode transitions. In this work, we extend classical Hamilton-Jacobi (HJ) reachability analysis, a formal verification method for continuous-time nonlinear systems, to hybrid dynamical systems. Our framework characterizes safe sets for hybrid systems through a generalized value function defined over both discrete and continuous states while accounting for control constraints and model uncertainty. We additionally provide a numerical
algorithm to compute this value function. Building on these safe sets, we propose two different mechanisms to integrate performance objectives. First, we introduce a hybrid least-restrictive safety filter that intervenes on both the discrete and continuous components of a nominal controller only when necessary to avoid unsafe states, thereby preserving nominal behavior whenever possible. Second, we formulate and compute hybrid backward reach-avoid tubes, enabling the simultaneous enforcement of safety and goal-reaching behavior, an extension not previously addressed within hybrid HJ reachability. This enables the synthesis of continuous and discrete control policies that guarantee both safety and task completion. We validate our framework through simulation studies and real-world experiments on a quadrupedal
robot, demonstrating its effectiveness in hybrid mode planning and safety-critical applications.
Read the full abstract ▾
Machine Learning
Learning-Guided Planning in Large Dynamic Action Spaces: Budgeted Tree Search for One-to-Many Mobile Charging
Liang-Ching Tao and Pi-Chung Wang
Many learned sequential decision systems map the current state directly to an action. That shortcut becomes brittle when candidate actions are numerous, geometrically structured, and rebuilt with the state. One-to-many mobile charging makes this setting concrete: with N=250 sensors, the initial state induces about 1,125 candidate charging-stop actions; each chosen stop simultaneously serves its in-range sensors, and the action universe changes as sensors die. LP-BTS is a learning-guided planning architecture: a graph proposal policy concentrates a small candidate support, a learned value critic evaluates leaves, and edge-budgeted PUCT compares short simulated futures before committing an action. Because the policy scores this set without a fixed output head, a single frozen checkpoint covers every evaluated setting, spanning action universes from 736 to 2,813 stops. Matched ablations reveal complementary effects: uniform sampling costs 8.8 survival percentage points, while, with targeted support fixed, PUCT jointly retains 1.4 points (about 3.5 of 250 sensors) and direct policy selection travels 23% farther. On a prospectively specified, sealed 30-scenario confirmatory bank evaluated once, LP-BTS attains the highest observed survival (0.4545) and alive-AUC (0.8031). Its estimated survival advantage over the strongest domain-engineered comparator is +0.0066 (95% CI [-0.0037, +0.0184]), an unresolved difference, while it exceeds a deadline heuristic and two source-derived direct-policy reconstructions on every paired scenario. Both learned rows are trained, source-derived reconstructions of variants reported by Gong et al. In this setting, the results provide controlled evidence about learning-guided planning in a large, dynamic action space.
Read the full abstract ▾