Computational Engineering
Topology optimization of force densities for form finding of cable structures
Nicolò Pollini
This paper presents a
topology optimization approach for the form finding of cable networks based on the force density method. The proposed framework combines the force density method with
topology optimization based on continuous density variables to identify the equilibrated geometry and effective connectivity of cable structures simultaneously. Design variables are assigned to the cable network members and control their force densities through a SIMP interpolation, complemented by an explicit binary promoting term in the objective function. As a result, cable members vanish or remain active throughout the optimization process. The
equilibrium configuration is obtained by enforcing the force density method equations, while the objective drives the optimized network as close as possible to a prescribed reference form. Volume constraints, passive boundary regions, and a minimum connectivity requirement at loaded joints are incorporated to yield physically meaningful and numerically stable designs. Several numerical examples based on cable net ground structures defined by horizontal, vertical, and diagonal members are used to study the proposed approach. A dedicated rounding study shows that the relaxed density variables converge to nearly discrete values at the end of the optimization process, with only a minor correction required after thresholding. A study of the volume fraction constraint further shows that reducing the permitted structural volume can trigger a qualitative transition in the optimized
topology, from a redundant, densely braced configuration to a minimal load path. Finally, an assessment of solver performance shows that the proposed formulation scales reasonably well with problem size.
Read the full abstract ▾
Information Retrieval
Tlow: Flow-based Item Tokenizer for Recommendation
Nian Li, Chonggang Song, Jingtao Ding et al. (+3 more)
Item tokenizer encodes semantic
embeddings into
token IDs to replace the randomly assigned item IDs used in traditional recommendation models, fundamentally addressing the problems of excessive parameters and cold starts. However, the most common tokenizer, RQ-VAE, suffers from low decoding efficiency due to the inherent dependencies among its codebooks. Meanwhile, efficient independent tokenizers such as optimized product quantization (OPQ) still struggle with dimensional correlations and distribution complexity of semantic
embeddings. In this work, we propose a f\underline{low}-based item \underline{T}okenizer (Tlow) to transform raw semantic
embeddings into a latent space where
embeddings conform to a unified standard normal distribution, achieving dual advantages of dimensional independence and distributional simplicity. Independent tokenization performed on these latent
embeddings yields semantically clear
token IDs. Additionally, we introduce a novel codebook guidance to align the codebook space with the
token embedding space, further aiding the learning of more semantically distinct
token embeddings. Offline experiments on four public
datasets demonstrate that Tlow's tokenization and codebook guidance significantly improve recommendation performance. The improvement on cross-domain and multi-modal recommendations also proves the effectiveness of item tokenization in a simplified
embedding space. Online experiments for a multi-modal
retrieval task on China's largest social media platform WeChat validate Tlow's powerful distribution transformation capability. The
retrieval model based on
token IDs improves user CTR by 10.32% globally and by 11.64% for new items. Our codes are available at https://github.com/wjjln/Tlow.
Read the full abstract ▾
All authors Nian Li, Chonggang Song, Jingtao Ding, Lingling Yi, Yong Li, Qingmin Liao
Computer Vision
Amortized Set Prediction for Inverse IFS Reconstruction from Density Maps
Yutaka Yamaguti
Iterated Function Systems (IFS) generate self-similar fractals from a few contractive affine maps. The forward map from parameters to images is inexpensive and well understood, whereas the inverse problem is difficult and is typically solved by per-image optimization. We replace this loop with a single forward pass of a learned
estimator that predicts the affine-map set directly from a visit-frequency density map, thereby amortizing the inverse problem. The design follows two constraints. First, density maps do not uniquely identify IFS parameters, so we evaluate by reconstruction rather than parameter recovery and handle unordered map sets by Hungarian matching. Training nevertheless matches the ground-truth parameters: reconstruction-only training gives a better one-shot output, but refinement from its predictions converges to poorer minima, whereas refinement from parameter-matched predictions keeps improving. Second, the fully known forward model lets us generate exact synthetic training pairs and supports image-only test-time refinement. On in-distribution tests, amortized initialization plus a few refinement steps gives a better quality-speed trade-off than equal-budget random-start optimization, and over 1000 steps it reaches high-quality reconstructions more often. It is robust to the sampling settings of the observation, and on an external
benchmark outside its training prior it matches the mean accuracy of a recent hour-long optimizer within a minute. On real images (MNIST and Fashion-MNIST), it improves density metrics on average over a published per-image optimizer while being roughly 2 to 500 times faster per image.
Read the full abstract ▾
Artificial Intelligence
Task-Adaptive Rubrics for GUI Reward Modeling
Tao Xiong, Xavier Hu, Wenkai Wang et al. (+6 more)
Recent studies on GUI
agents have increasingly focused on outcome reward modeling, which assigns outcome rewards by judging whether an executed trajectory satisfies the success criteria implied by the user instruction. Existing GUI reward verifiers, however, often under-specify how these criteria should be constructed for each task instance. Whether using generic rubric structures or implicit model
reasoning, their judging criteria are not sufficiently task-adaptive: they can transfer checks across tasks, overlook concrete constraints in the current instruction, or become overly strict by enforcing unstated requirements. To address this limitation, we propose AdaptRubric, a Coarse-to-Fine Rubrics Framework that constructs task-adaptive judging criteria through a category-level coarse stage and an instance-level fine stage. AdaptRubric performs category-level coarse rubric
retrieval by routing the instruction to a GUI task family and retrieving reusable task-family criteria, then conducts instance-level fine rubric generation to surface compact cues for concrete values, scopes, and constraints in the current instruction. Across offline reward evaluation and online reinforcement learning optimization, AdaptRubric consistently outperforms prior reward
agents, improving F1 by 3.6 points over the baseline average under a matched image budget and yielding a 4.23-point task-success gain.
Read the full abstract ▾
All authors Tao Xiong, Xavier Hu, Wenkai Wang, Qinzhuo Wu, Changqiao Wu, Pengzhi Gao, Wei Liu, Jian Luan, Shengyu Zhang
Computer Vision
SandwichQuant: Which Parameters Matter Before and After Quantization?
Peng Xia and Junbiao Pang
Quantization correction methods usually optimize weights, quantization parameters, or reconstruction objectives, while the underlying parameter subspaces responsible for effective correction remain unclear. In this work, we study quantization correction from a parameter subspace perspective and reveal that correction capability is highly non-uniform across parameter groups. By decomposing trainable parameters into backbone weights, normalization-affine parameters, and quantization parameters, we show that the low-dimensional normalization-affine subspace provides a highly efficient correction direction under matched budgets. Based on this finding, we propose SandwichQuant, a two-stage normalization-affine correction framework that performs adaptation before and after quantization. The pre-stage improves quantization robustness, while the post-stage compensates residual errors after the quantized graph is fixed. Extensive experiments on
vision models and large
language models demonstrate consistent improvements under various low-bit quantization settings, validating the effectiveness of subspace-aligned correction.
Read the full abstract ▾
Computer Vision
ViSculpt: Visual-Centric Agentic Geometry Editing
Bo Pang, Jiaqi Pan, Xiaocheng Zhang et al. (+3 more)
3D geometry editing is a critical yet labor-intensive part of the graphics pipeline, requiring artists to translate creative intent into precise operations in complex professional software. Large
language models (LLMs) have shown promise for script-based 3D creation, but script generation is less suited to perception-driven editing of arbitrary existing meshes, where execution must remain visually grounded and untouched regions should be preserved. We present a visual-centric, training-free multi-
agent system that edits existing 3D meshes directly in Blender by emulating the iterative workflow of human artists. Rather than generating scripts or regenerating geometry, our system operates through the Blender GUI:
multimodal LLM
agents observe the viewport, reason about the current mesh state, and execute localized edits through simulated user interactions. Experiments on a curated
benchmark provide initial evidence that this agentic approach can follow natural
language instructions, perform representative localized mesh edits, and preserve the overall identity of the input asset. Our results highlight a complementary regime for
language-driven 3D editing: direct in-place modification of existing meshes within the native 3D editing workflow. We view this work as an exploratory step toward visual-centric agentic geometry editing in professional graphics software.
Read the full abstract ▾
All authors Bo Pang, Jiaqi Pan, Xiaocheng Zhang, Jiacheng Xu, Guoping Wang, Peng-Shuai Wang
Computation and Language
FireRedAudio: A General-Purpose Audio Language Model with Decoupled Continuous Representations for Understanding and Generation
Feiyu Shen, Fenglong Xie, Junjie Li et al. (+13 more)
A unified audio model must recognize and understand linguistic, paralinguistic, and environmental information while supporting
speech synthesis and editing. A key challenge is representation: understanding favors compact features suited to long-context modeling, whereas
speech generation requires reconstructible features that preserve fine-grained acoustic detail. We introduce FireRedAudio, a general-purpose audio
language model with a shared 9B-parameter LLM. To the best of our knowledge, it is the first publicly disclosed unified audio-
language model to provide separate continuous input representations for understanding and generation within a single trainable autoregressive LLM. Audio to be recognized or analyzed is processed by a dedicated Audio Encoder, while
speech inputs for generation use a RedAE-based pathway. The LLM directly generates text or conditions a flow-matching DiT to produce continuous acoustic latents. Through progressive multitask training, FireRedAudio supports ASR and audio understanding, with the latter extending to recordings of up to one hour, as well as zero-shot TTS, Instruct TTS, and semantic and acoustic
speech editing. Its structured organization of long-form audio achieves second-level timestamp accuracy. Across comprehensive evaluations, FireRedAudio achieves competitive or leading performance in audio understanding and
multilingual ASR, strong content accuracy and speaker preservation in zero-shot TTS, leading instruction following in Instruct TTS, and substantial improvements over Ming-UniAudio-Edit in both semantic and acoustic
speech editing. These results demonstrate the viability of decoupled continuous input representations for unifying audio understanding and continuous-latent
speech generation in a model of moderate scale. Our code is available at https://github.com/FireRedTeam/FireRedAudio.
Read the full abstract ▾
All authors Feiyu Shen, Fenglong Xie, Junjie Li, Kun Xie, Lei Xie, Xu Tang, Xuelong Geng, Yan Jia, Yao Hu, Yichen Han, Yichen Wu, Ziqi Dai, Junjie Chen, Kai Huang, Manzhen Wei, Yixuan Li
Machine Learning
From Relaxed Indexability to Exact Indexability: A $t$-Step Approach for Partially Observable Restless Bandits
Keqin Liu and Qizhen Jia
Whittle index policies offer a scalable method for restless multi-armed bandits, but under partial observability even determining the indifference subsidy at a single belief requires solving an infinite-horizon belief-state problem with no closed-form value function. Liu [10] addresses this difficulty by linearizing the unknown decision boundary, leading to a linear system and a closed-form approximate Whittle index. However, the resulting threshold uses only a one-step active–passive comparison and does not account for longer-horizon continuation values. We extend this framework to a \emph{$t$-step lookahead threshold policy}. For each subsidy $m$, the threshold is defined by the active-minus-passive advantage under $t$-step finite-horizon value iteration. At $t=1$, the threshold is $m$-independent and recovers the linear threshold of Liu [10]; for $t>1$, it becomes subsidy-dependent through the induced first-crossing structure and tracks the exact decision boundary more closely. The proposed algorithmic framework does not require indexability as an input and includes an indexability verification. Under the original Whittle indexability, we prove that the $t$-step approximate Whittle index converges geometrically to the exact Whittle index, \[ |\widehat W_t(\omega)-W(\omega)|=O(\beta^t). \] Numerically, all 2,715 tested three-state instances are verified with computable priority index functions according to the proposed criterion. The P95 index error decreases from $2.18\times10^{-2}$ at $t=1$ to $8.93\times10^{-4}$ at $t=8$. In an exact-comparable instance with $\beta=0.9999$, $t=2$ already recovers the exact Whittle-index ordering. Moderate-depth threshold policies also outperform the one-step baseline and remain close to the optimal dynamic-programming
benchmark, while runtime grows mildly with $t$.
Read the full abstract ▾
Computer Vision · Cross-listed from astro-ph.GA
Decoupling candidate dual AGN from chance superpositions in the GOTHIC survey via a deep-learning framework
Bhavesh Mukheja, Snehanshu Saha, Anwesh Bhattacharya et al. (+3 more)
Dual active
galactic nuclei (DAGN) mark a critical phase in the
evolution of merging
galaxies and the pairing of supermassive black holes, yet they remain difficult to identify in large imaging surveys because of projection effects and limited spatial resolution. Compact foreground
stars and unresolved substructure can mimic dual
nuclei through chance superposition, complicating automated detection. We revisit the 46,061
galaxies flagged but rejected as DAGN candidates by the GOTHIC pipeline, primarily because the two
nuclei fell within the SDSS fibre aperture or exceeded its separation threshold. We train a supervised deep-learning framework based on the YOLOv11 oriented-bounding-box architecture on annotated SDSS imaging to separate genuine dual
nuclei from foreground
stellar contaminants and other spurious
alignments. The final model attains a validation precision of 0.919, recall of 0.905, and $F_1$ of 0.912, and yields 29,605 dual-
nucleus candidates after removing
star-dominated and blended detections. Structured visual inspection indicates that $54.5$–$62\%$ are consistent with genuine dual
nuclei, implying $\sim(1.4$–$1.8)\times10^{4}$ plausible systems. Cross-calibrating the YOLO separation against the deterministic GOTHIC centroid measurement and restricting to the compact regime ($d \le 6.87''$) gives a conservative subset of $\sim 13{,}672$ candidates, reaching calibrated separations of $\sim 0.56''$. Spectroscopy of the most compact ($\le 1$ kpc) systems shows spectra dominated by
stellar absorption, with no resolved double-peaked emission, so confirmation requires higher-resolution follow-up. The catalogue is a statistically refined list of candidates, not confirmed DAGN. Nonetheless, deep-learning detection substantially reduces contamination and expands the plausible DAGN census.
Read the full abstract ▾
All authors Bhavesh Mukheja, Snehanshu Saha, Anwesh Bhattacharya, Mousumi Das, Françoise Combes, Sudhanshu Barway
Artificial Intelligence · Cross-listed from eess.AS
Preference Optimization for Non-Verbal Vocalization Synthesis
Haoyang Li, Chenglin Xu, Junchuan Zhao et al. (+4 more)
Non-verbal vocalizations (NVs), such as laughter, coughs, and sighs, are essential for expressive TTS, but the effectiveness of preference optimization for NV generation remains poorly understood. We systematically study preference optimization for NV-capable TTS, focusing on preference signals, preference-pair construction, and DPO-based optimization objectives. We formulate an NV-aware character error rate (NV-CER) by treating NV tags as distinct output symbols and computing a weighted pinyin-based CER over both verbal and non-verbal content, enabling controllable optimization of NV realization without modifying the underlying optimization
algorithm. Experiments on Emilia-NV and the augmented NV-Bench covering 18 NV types reveal how different design choices affect NV realization and lexical fidelity, and establish an effective setup using standard DPO. Objective, LLM-based, and human evaluations provide converging evidence for our findings, offering practical insights into NV-aware post-training for expressive TTS.
Read the full abstract ▾
All authors Haoyang Li, Chenglin Xu, Junchuan Zhao, Yuang Cao, Liumeng Xue, Yiwen Guo, Eng Siong Chng
Robotics
Robust Slip Detection and Material Classification via Spatiotemporal Transformers on a Uniformly-Illuminated Visuo-Tactile Sensor
Ziyang Ma, Yuhao Sun, Zichen Ai et al. (+2 more)
Tactile sensing is central to robotic manipulation, among which slip detection stands out as a quintessential and critical task. However, existing slip
datasets are predominantly limited to binary classification, lacking fine-grained directional perception. To address this limitation, we propose a visuo-tactile sensor featuring customized uniform RGB illumination, alongside a unified perception framework. At the hardware level, the sensor achieves high-precision, sub-millimeter depth reconstruction. Based on this capability, we collect a multi-task visuo-tactile
dataset encompassing 15 objects, synchronously generating depth information for each data sample. Algorithmically, we design a dual-head TimeSformer network to process dynamic spatiotemporal slip. On unseen objects, this network achieves robust accuracies of 95.5% and 91.5% for 3-class contact state prediction and fine-grained 8-class slip direction classification, respectively. Furthermore, static tactile-based object class recognition utilizing a ResNet-50 backbone yields an outstanding accuracy of 98.8% across 15 categories. The proposed hardware-software framework provides high-fidelity feedback and a powerful multi-modal perception baseline for complex robotic manipulation.
Read the full abstract ▾
All authors Ziyang Ma, Yuhao Sun, Zichen Ai, Xiangyang Ji, Bin Fang
Artificial Intelligence
OmniJudge or OmniBias? Diagnosing Multimodal Judges through Balanced, Decoupled Lenses
Guangzheng Hu, Ziyue Jiang, Weixu Qiao et al. (+14 more)
Multimodal understanding models that can jointly judge text-to-image (T2I), text-to-
video (T2V) and text-to-
speech (TTS) generation are increasingly used as "OmniJudges" for evaluation and automatic annotation. How reliably they understand what they score remains unclear, since existing
benchmarks and training data tend to overemphasize positive examples and to conflate distinct failure modes, so a judge may score well without recognizing failures while its capability gaps stay hidden. Motivated by this, we introduce D3-Omni, a balanced and decoupled
benchmark for diagnosing fine-grained
multimodal understanding, covering 53 orthogonal binary dimensions (17/22/14) and 10,671 samples (3,526/1,998/5,147) across the three tasks. Rather than re-generating outputs, which may leak information across dimensions, we fix verified fully positive seeds and derive negatives through controlled
prompt rewriting and atomic, dimension-isolating perturbations. The resulting D3 design is Dual-balanced, which helps alleviate negative-sample scarcity and per-dimension label imbalance; Decoupled, so that each error is attributable to a single capability; and Dynamic, steering construction toward under-represented regions of the label distribution as generative models improve.The suite reaches near 1:1 per-dimension parity and a uniform distribution over all total-score levels. Under this balanced view, even strong OmniJudges tend to struggle on modality-related dimensions, to confirm satisfied requirements far more reliably than they detect violated ones, and to treat nominally distinct attributes as largely a single decision, suggesting that aggregate accuracy may hide systematic blind spots that a balanced and decoupled lens can help expose and, in turn, address.
Read the full abstract ▾
All authors Guangzheng Hu, Ziyue Jiang, Weixu Qiao, Lixin Zhang, Jianye Kang, Yuru Wu, Rong Bao, Niantong Li, Wei Wang, Ziyi Cheng, Xinfa Zhu, HangRui Hu, Ting He, Bing Zhao, Lin Qu, Hu Wei, Jin Xu
Sound
On the Robustness of Audio Deepfake Detection under Audio Watermarking
Zi Qian Yong, Ajinkya Kulkarni, Julia Lau et al. (+4 more)
Recent advances in generative audio models have enabled highly realistic synthetic
speech, increasing the importance of reliable audio deepfake detection (ADD) systems. While prior studies have primarily focused on adversarially optimized perturbations, the robustness of ADD systems under realistic signal transformations remains insufficiently understood. In this work, we investigate the impact of audio watermarking on ADD systems by treating watermarking as a structured, non-
adversarial perturbation rather than a conventional attack mechanism. Using a watermark-based evaluation framework built upon WavMark, we evaluate multiple self-supervised learning (SSL), Convolutional Neural Network (CNN) and Graph Neural Netrowk (GNN)-based ADD models across several
benchmark datasets. Beyond conventional detection metrics, we further analyze watermark-induced representation shifts using Fréchet Distance, cosine similarity, and L2 distance in the
embedding space. Experimental results reveal a strong
dataset-dependent behavior: watermarking causes substantial performance degradation on ASVspoof 2021 LA and DF, while exhibiting limited impact on ASVspoof 2024, FoR, and ITW. Moreover, large
embedding-space shifts are strongly associated with severe detection degradation, suggesting that watermark-induced perturbations can substantially alter the feature representations relied upon by current ADD systems. These findings demonstrate that benign signal transformations designed for content protection can expose previously overlooked robustness vulnerabilities in audio deepfake detection systems. Our code is available at https://github.com/ziqian0925/wm-ADD-robustness.git
Read the full abstract ▾
All authors Zi Qian Yong, Ajinkya Kulkarni, Julia Lau, Hwa Hui Tew, Shu Min Leong, Raphael Phan, Sébastien Marcel
Human-Computer Interaction
Balancing Evidence and Interpretation: Historical Grounding Ratio as a Design Parameter for AI-Generated Urban Storytelling
Fuyang Zhang and Maurice Benayoun
Location-aware generative systems can now select historical archives and real-time contextual information based on a user's surroundings to automatically generate narratives for urban heritage walks. Yet when multiple sources jointly inform generation, existing systems provide neither a clear representation of how much content from each source actually appears in the output nor an operational means of measuring it. We introduce the Historical Grounding Ratio (HGR), defined as the proportion of claim-bearing information units in a generated narrative that are supported by historical archives. HGR turns the realized share of historical evidence in a narrative into a directly measurable design parameter. In GeoDrama, a mobile narrative system, we created three conditions that used a common
retrieval procedure and comparable evidence-bundle sizes while varying the allocation of information from different sources during generation. We evaluated how changes in HGR affected narrative experience through a within-subject walking study with 18 participants. Increasing HGR significantly strengthened the perceived relevance between narrative content and the specific location. However, historical understanding, integration with the visible scene, appropriateness of the amount of information, and intention to explore further did not increase monotonically with HGR; all four measures were highest in the intermediate, balanced condition. These findings show that designing location-aware generative interfaces involves not only retrieving relevant material but also determining how information from different sources composes the final output. HGR offers an operational measure for comparing information-allocation strategies and their experiential consequences.
Read the full abstract ▾
Systems and Control
LLM-Guided Contextual Action Evaluation for Operational Decisions in Industrial Processes
Youcheng Zong, Runda Jia and Dakuo He
Industrial actor–critic methods usually represent continuous actions as anonymous numerical coordinates. They must therefore learn from limited interactions which process variables each action affects, in which direction, and after what delay. Fixed industrial documents already describe part of these relations, but their open-text statements neither represent the current operating condition nor directly fit a numerical policy. This article presents LLM-Guided Contextual Action Evaluation for Operational Decisions in Industrial Processes (LCAE), which uses a large
language model before training to normalize fixed documents into a frozen action–observation–direction–delay relation basis. Recent numerical action–response history then modulates the current strength of each relation, while the evaluated action forms a state-conditioned nonlinear action-effect field in the same basis. The critic evaluates actions through this field, and the actor uses the same relation gains to generate actions, making document semantics part of maximum-
entropy policy learning. Neither the LLM nor the
embedding model runs online during training or deployment; the deployed policy uses only frozen semantic artifacts and visible numerical history. The method states a falsifiable hypothesis: when documented relations are correct and recent history reflects their contextual strength, this action representation should provide a more useful decision bias than raw action coordinates.
Read the full abstract ▾
Robotics
Coverage Planning for Robotic Tooth Preparation in Densely Constrained Environments
Yunwen Li, Chen Chen, Xiangjie Yan et al. (+4 more)
Tooth preparation refers to the controlled removal of tooth structure to create an optimal substrate for fixed restorations and is a core procedure in restorative dentistry. Automating this task is particularly challenging for
robots because the dental bur must operate within a densely constrained intraoral workspace, where even sub-millimeter deviations can compromise outcomes or damage adjacent structures. This paper presents a novel robotic system for autonomous full-crown tooth preparation. The proposed framework includes: 1) an anatomy-aware toolpath planning
algorithm that conforms precisely to a technician-designed preparation model while protecting adjacent teeth, and 2) a clearance-oriented end-effector yaw assignment strategy that allows intraoral access while reducing the risk of soft-tissue interference. Together, these features enable the
robot to accurately mill the irregular tooth surface with an average
geometric deviation of 0.117 mm (RMSE), achieving both restoration quality and clinical safety. A series of simulations and phantom-head experiments validate the system's feasibility and effectiveness.
Read the full abstract ▾
All authors Yunwen Li, Chen Chen, Xiangjie Yan, Chang Shu, Jianxia Hou, Shiji Song, Xiang Li
Computer Vision
Rethinking Pre-Training and Augmentation for Zero-Shot Cross-City Object Detection
Long Hoang Pham, Quoc Pham-Nam Ho, Huy-Hung Nguyen et al. (+10 more)
Real-world deployment of traffic surveillance systems is bottlenecked by geographic domain shift, in which models trained in one city underperform when applied to an unseen target city. Conventional domain adaptation relies on hyperparameter-sensitive architectures or direct profiling of target data. Both are fundamentally precluded in
privacy-conscious ecosystems that require completely blind training and evaluation loops. In this setting, we explore the effects of pre-training and augmentation in addressing the domain shift problem. Specifically, we propose a new modular training pipeline for object detection structured around two core orthogonal pillars: (1) a multi-
dataset pre-training strategy featuring a class-agnostic objectness
distillation to decouple structural vehicle geometry from semantic taxonomies, and (2) a domain-resilient augmentation stream featuring a novel Grayworld transformation that forces global
attention heads to strip volatile chromatic shortcuts in favor of robust shape priors. When evaluated with the real-time
transformer-based detector RF-DETR, our framework bridges cross-city distribution gaps while using limited GPU
memory (16GB). Our optimized variants, RF-DETR-HR and RF-DETR-Grayworld, deliver a substantial empirical gain of +24.29 over the baseline, achieving 1st place (47.53 mAP) on the AI City Challenge Track 6 leaderboard. Code and data are available at: \href{https://github.com/SKKUAutoLab/aic26_cross_city}{SKKUAutoLab/aic26\_cross\_city}.
Read the full abstract ▾
All authors Long Hoang Pham, Quoc Pham-Nam Ho, Huy-Hung Nguyen, Duong Nguyen-Ngoc Tran, Ngoc Doan-Minh Huynh, Cu Quoc Le, Hoang-Khang Nguyen, Hyung-Min Jeon, Chi Dai Tran, Son Hong Phan, Duong Khac Vu, Trinh Le Ba Khanh, Jae Wook Jeon
Distributed Computing
pigzpp: Fast, Parallel, Portable Compression for the Whole Stack
Thamme Gowda
pigz is a widely deployed parallel gzip utility, but its process-global mutable state means that it was not designed as a reentrant, directly embeddable library. pigzpp is a from-scratch C++23 rewrite that turns the design into a thread-safe library with one accelerated DEFLATE core exposed to C++, Python, WebAssembly, Go, and Rust. As application-level conveniences built on that same core, it also provides multi-entry ZIP archives and a fast PNG codec. Its portable zlib-ng backend preserves gzip/zlib's ratio; its Intel ISA-L backend is an x86-64-only fast path that produces about 10% larger output at level 6 on our text corpus. In that configuration, zlib-ng reaches 2.1 times the CLI throughput of pigz and 16 times that of Python's in-
memory gzip; ISA-L reaches 8.5 times and 50 times, respectively. Outputs remain standards-compliant and cross-decode with gzip, pigz, and unzip. The result is a self-contained, multi-platform compression stack and a case study in AI-assisted modernization under automated compatibility tests.
Read the full abstract ▾
Networking · Cross-listed from quant-ph
A Dynamic-Kernel/QPacket Executable for Quantum Repeater Chains in Q2NS/ns-3
Adam Pearson, Marcello Caleffi and Angela Sara Cacciapuoti
The
Quantum Internet operates on
entanglement, a non-local, non-copyable, stateful network resource, which motivates protocol organization beyond classical layering. We present a first executable specialization of the Dynamic Kernel/QPacket logic from the beyond-layering protocol suite, targeting
entanglement distribution over a linear
quantum repeater chain. The implementation builds on Q2NS, an ns-3-based
quantum-network simulation module available through the ns-3 App Store. It realizes QPacket meta-headers with service intent and append-only action-commit stamps processed by node-local Dynamic Kernels organized as a Planner–Executor–Engine pipeline, while being deliberately scoped to an analytically verifiable service and policy. Within this scoped setting, we study node heterogeneity through a link-preparation policy that accounts for pre-distributed
entanglement and uneven
entanglement-generation support across nodes, including delegation via QPacket forwarding. Simulations verify analytical link-resolvability models and expose signaling load, forwarding behavior, and QPacket meta-header growth. Results show that QPacket overhead is shaped by more than just encoding, including policy choices and available network resources. Overall, this study demonstrates how the Q2NS/ns-3 substrate can support reproducible, policy-specific evaluation of
quantum-native protocol-suite concepts.
Read the full abstract ▾
Numerical Analysis
Condition numbers of block Toeplitz matrices and stability of space-time IgA approximations for the wave and Schrödinger equations
Manuel Bogoya, Albrecht Böttcher, Matteo Ferrari et al. (+2 more)
In previous work by several authors, the behavior of the condition numbers of banded Toeplitz
matrices was studied as the
matrix size tends to infinity. In the present contribution, two main directions are pursued. As a first step, we extend this study to block Toeplitz
matrices with blocks of fixed size $N$. As in the scalar case, we show that even when the symbol generates a Fredholm infinite Toeplitz operator, the condition numbers of the finite
matrices may grow at least exponentially. Upper and lower bounds for the condition numbers are obtained, and examples showing that they may grow arbitrarily fast are presented. Then, as a second step, we apply the developed theory to the stability analysis of space-time Galerkin methods, where in time an Isogeometric approach is used with regularity $r$, $1\le r\le p-1$, $p$ being the employed
polynomial degree. These stability issues are related exactly to the conditioning of block Toeplitz-like
matrices with blocks of size $N=p-r$. Specific examples are treated in detail and related numerical experiments are presented and critically discussed. We finally present a short list of relevant open problems.
Read the full abstract ▾
All authors Manuel Bogoya, Albrecht Böttcher, Matteo Ferrari, Sergei M Grudsky, Stefano Serra-Capizzano
Machine Learning
Robust Data-Collection Policy Learning for Low-Variance Online Policy Evaluation
Claire Chen, Shuze Daniel Liu, Licheng Luo et al. (+3 more)
In reinforcement learning policy evaluation, classic on-policy methods often suffer from high variance when estimating policy performance. To mitigate this issue, behavior policy search has been proposed to learn data-collecting policies tailored to reduce online evaluation variance. However, these approaches do not account for uncertainties in the transition functions. In practice, simulator transitions often differ from the real world due to modeling errors or approximation limitations. As a result, behavior policies trained in simulation may still yield high variance when deployed in real environments, leading to costly reliance on real-world evaluation samples. In this work, we propose a double-loop gradient-based
algorithm for learning behavior policies that are both efficient and robust to transition uncertainty. Theoretically, we derive novel transition-variance gradient expressions and establish global convergence guarantees for the
algorithm. Numerically, we demonstrate that our method is less sensitive to transition perturbations than existing approaches, providing supportive evidence for its practical utility.
Read the full abstract ▾
All authors Claire Chen, Shuze Daniel Liu, Licheng Luo, Rohan Chandra, Nan Jiang, Shangtong Zhang
Computation and Language
TrustDABench: Benchmarking Reliability and Robustness of LLMs for Structured Data Analysis
Boshen Shi, Yize Liu, Chen Zhao et al. (+4 more)
LLMs are increasingly used to analyze spreadsheets, CSV files, and other structured data, but producing a correct-looking answer is not the same as producing a trustworthy analysis. A trustworthy result should be supported by a valid path from the user question to the relevant data evidence. This requirement creates two diagnostic questions: whether an LLM can refuse to answer or ask for clarification when such a path does not exist, and whether it can preserve the correct analysis when the same evidence is expressed in different table forms. We introduce TrustDABench, a
benchmark that operationalizes these questions as reliability and robustness. Starting from the evidence-path view, we derive 19 perturbation operators and instantiate them through an Agentic-LLM-based generation framework. TrustDABench contains 2,340 human-verified perturbed instances, and we evaluate eight representative LLMs. The results show substantial headroom: the best reliability result is only 24.21% average MRS, achieved by GPT-5.5, while the best robustness result still has 9.10% average ASR, achieved by Claude-Sonnet-5. The failures are systematic: models rarely detect conflicting evidence, often continue along executable but unsupported analysis paths, and remain sensitive to perturbations that change observation boundaries or cross-table relations. These findings suggest that stronger evidence-boundary recognition and representation-invariant
reasoning are still needed for reliable structured-data analysis.
Read the full abstract ▾
All authors Boshen Shi, Yize Liu, Chen Zhao, Ce Chi, Zhendong Wang, Xing Wang, Junlan Feng
Symbolic Computation
StabQ: Quantum Program Analysis via Weighted Stabilizer Representations
Shangzhou Xia, Junjie Luo and Jianjun Zhao
Quantum program analysis remains challenging due to the exponentially large state space of
quantum programs and the difficulty of precisely characterizing their execution behavior. In particular, non-Clifford operations introduce additional complexity that limits the applicability of stabilizer-based techniques. Although stabilizer representations provide compact descriptions for Clifford
circuits, their limited expressiveness prevents them from directly supporting general
quantum program analysis. In this work, we propose StabQ, a symbolic execution framework for
quantum program analysis based on stabilizer representations. StabQ extends stabilizer-based symbolic execution beyond Clifford-only programs by introducing a symbolic state representation that captures and propagates
quantum state
evolution while preserving execution semantics. Based on this representation, StabQ constructs a Tableau Chain that represents the
evolution of intermediate symbolic states throughout program execution and enables reusable analysis of
quantum program executions. Furthermore, StabQ incorporates tableau consolidation and global-phase recovery mechanisms to mitigate symbolic state growth during execution. Building upon the Tableau Chain, StabQ supports multiple
quantum program analysis tasks, including
quantum state reconstruction,
entanglement analysis, and Clifford-property detection. We evaluate StabQ on three
benchmark suites—
Algorithms, MQT Bench, and QASMBench. The results demonstrate that StabQ constructs semantically consistent symbolic models, accurately preserves
quantum state
evolution, and effectively supports downstream analysis tasks across diverse
quantum programs.
Read the full abstract ▾
Computer Vision
What Does Prompt Learning Change? -A Natural-Language Concept Analysis of Vision-Language Models
Ryo Kamiya, Hiroshi Kera and Kazuhiko Kawamoto
Prompt learning adapts
vision-
language models such as CLIP by optimizing continuous
prompt vectors, but the learned
prompts are difficult to interpret in natural
language. We present PromptSpLiCE, a post-hoc method that expresses each class-conditioned text
embedding as a sparse combination of concepts from a fixed natural-
language dictionary. Using the same dictionary before and after
prompt learning allows us to compare changes in their concept profiles. We evaluate PromptSpLiCE on CoOp, a representative
prompt-learning method, across 11 image-classification
datasets. The concept profiles change substantially: on average, only 1.6 of the initial top-10 concepts remain in the top 10 after learning. Across
datasets, profile change is positively associated with accuracy gain. We also derive a local gradient expression that provides
geometric intuition for why image-aligned concept directions distinct from the current
prompt can have greater loss sensitivity.
Read the full abstract ▾
Computer Vision
Rubrics as Visual-Repair Context for Self-Evolving UI-to-Code Generation
Tianyi Xiong, Zhengyuan Yang, Xiaofei Wang et al. (+10 more)
Large
vision-
language models have shown strong progress in UI-to-code generation, yet their test-time self-
evolution remains unstable. We first identify a fundamental obstacle, termed visual repair coupling: a local code edit may propagate through layout, style, and component dependencies, correcting one visual mismatch while degrading regions that were previously faithful. To address this issue, we present RubSE, a Rubric-guided Self-
Evolution framework that uses rubrics to represent visual feedback as a structured visual-repair context. At each refinement round, RubSE generates typed candidate rubrics, selects one prioritized repair target, and stores previously selected rubrics as history, thereby steering each revision toward a well-scoped visual repair while discouraging repeated or over-broad changes. Evaluations across six VLMs and three UI-to-code
benchmarks demonstrate that RubSE substantially outperforms naïve self-
evolution in final-round and best-round settings, achieving more stable refinement trajectories and a higher trajectory-level performance ceiling. Further analysis shows that RubSE mitigates trajectory collapse by improving recovery from severe visual regressions, and that stronger rubric generators can transfer effective visual-repair guidance to weaker code improvers.
Read the full abstract ▾
All authors Tianyi Xiong, Zhengyuan Yang, Xiaofei Wang, Chung-Ching Lin, Ruichun Ma, Kevin Lin, Zhendong Wang, Linjie Li, Chenxi Liu, Ruibo Chen, Ramani Duraiswami, Heng Huang, Lijuan Wang
Machine Learning
Steering Recurrent Reasoners at Inference Time with Readout Feedback
Shunsuke Kamiya, Masanori Koyama, Seongcheol Jeong et al. (+5 more)
Recurrent models, which repeatedly update latent states with shared computation blocks, have emerged as powerful architectures for solving complex
reasoning tasks. Existing inference-time methods scale computation by running more steps or sampling more trajectories, but ignore information revealed within each trajectory. Here we show that recurrent models can be improved at inference time by using their own readout probabilities to steer latent dynamics without retraining. We introduce Readout Feedback (RoFB), a test-time intervention that converts intermediate predictions into
token-wise pairwise coupling forces injected into the latent dynamics. Across three recurrent models (AKOrN, ItrSA++, TRM) on Sudoku and Maze, RoFB yields clear gains in four of six model-task pairs and a small gain in one. The four positive pairs achieve performance unattainable by merely running more steps or selecting from multiple trajectories, at comparable or lower computational cost. These results suggest that closed-loop steering of latent dynamics can serve as a complementary inference-time control mechanism for recurrent
reasoning models.
Read the full abstract ▾
All authors Shunsuke Kamiya, Masanori Koyama, Seongcheol Jeong, Fumiya Uchiyama, Kenji Kubo, Kohei Hayashi, Masahiro Suzuki, Yutaka Matsuo
Artificial Intelligence
Robust Code RL via Faulty-Code-Driven Test case Synthesis and Dense Reward Shaping
Yiwen Zhang, Xiaodong Yan, Zhenyu Huang et al. (+6 more)
Reinforcement Learning from Verifiable Rewards (RLVR) is pivotal for enhancing LLM code generation, yet its efficacy is often hindered by insufficient test case coverage, leading to reward hacking and policy degradation. To address this, we propose RobustTests, a framework featuring a faulty-code-driven test case synthesis strategy. By leveraging "near-correct" faulty codes, RobustTests captures latent logical discrepancies and employs validator
agents with behavioral feature clustering to filter invalid or redundant test cases. Additionally, a stepwise dense reward function based on pass rates is introduced to mitigate false negatives and enhance training robustness. Using this pipeline, we construct an augmented version of the CodeContests+
dataset with superior diagnostic utility. Experimental results show that RL fine-tuning of Qwen3-32B via RobustTests achieves a 3% absolute gain on LiveCodeBench, demonstrating its effectiveness in advancing LLM code generation proficiency. Codes and data are available at https://huggingface.co/
datasets/sid6/RobustTests.
Read the full abstract ▾
All authors Yiwen Zhang, Xiaodong Yan, Zhenyu Huang, Deng Zhao, Liang Jiang, Qing Cui, Zujie Wen, Zhiqiang Zhang, Jun Zhou
Computer Vision
EgoErrorVQA: Assess Egocentric Comprehension Capabilities through Procedural Errors for Ego-Agentic AI
Junlong Li, Junxi Li, Jianjun Gao et al. (+3 more)
The majority of our everyday activities are procedural and consist of sequences of interdependent steps. However, existing
benchmarks for Visual
Agents and Visual
Language Models (VLMs) overlook the evaluation of their procedural comprehension ability from an
egocentric visual perspective, particularly for detecting procedural errors, a critical capability for everyday assistance. To bridge this gap, the EgoErrorVQA task is firstly proposed for
egocentric procedural comprehension with explicit procedural errors modeling. Besides, we develop a user-friendly evaluator
agent based on the
Agent2
Agent (A2A) protocol, enabling rigorous and standardized evaluation of visual
agents through VQA-based interaction. A range of models are evaluated using both open-ended and multiple-choice questions, revealing persistent weaknesses in handling procedural errors and error types. Moreover, we introduce Ego-ADR, an Adaptive Decoupled
Reasoning framework that decouples complex procedural
reasoning to enhance models' understanding of procedural errors. It achieves consistent performance gains over the selected baselines and attains state-of-the-art results on several metrics under comparable settings. Code: https://github.com/z1oong/EgoErrorVQA
Read the full abstract ▾
All authors Junlong Li, Junxi Li, Jianjun Gao, Chen Cai, Lap-Pui Chau, Yi Wang
Computer Vision
PlaceSeek: Human-Centered Geospatial Retrieval of Urban Outdoor Places via Semantic Grounding and Affective Alignment
Ziqi Cui and Shangyu Lou
People search for urban outdoor places not only by category or function, but also by what activities a place can support and how it is perceived. Existing geospatial
retrieval remains largely POIcentric and metadata-driven, making it difficult to satisfy openended, affective, or activity-oriented needs. We present PlaceSeek, a human-centered outdoor place
retrieval framework that maps natural-
language queries to geolocated street-view imagery. PlaceSeek introduces an intent-aware
retrieval mechanism that decomposes user queries into functional and affective sub-intents. A Semantic Grounding Module verifies whether candidate street-view results contain the physical evidence needed to support the intended activity, while an Affective
Alignment Module re-ranks physically valid candidates using a LoRA-adapted
vision-
language model trained on human urban perception judgments. We evaluate PlaceSeek on 31,956 street-view locations in Milan across 10 naturallanguage queries annotated by five human evaluators. PlaceSeek achieves 88.0% Precision@5, a mean match score of 3.39/4.0, and 0.920 nDCG@5, outperforming CLIP, fine-tuned CLIP, SigLIP, and a VQA-based baseline. Ablation results show that physical grounding is essential for
retrieval validity, while affective
alignment improves ranking quality among physically valid candidates. These findings highlight that complex urban spatial queries require modeling both verifiable visual evidence and human perceptual preferences. PlaceSeek provides a potential framework for human-centered nextgeneration geospatial
retrieval systems.
Read the full abstract ▾
Machine Learning
From Gradient-Boosted Trees to Deep Recommenders: Practical Lessons from Migrating a Production Customer Support Recommender
Sonia Sharma, Jeyendran Balakrishnan, Shreya Rajpal et al. (+3 more)
Product catalogs in fast-moving service businesses are shifting from static, independently priced SKUs toward dynamically bundled, discount-coupled offerings–a shift that strains the tree-based classifiers traditionally preferred for sparse and highly imbalanced data. These classifiers assume a fixed, slowly changing label space and struggle to incorporate
multimodal signals such as tabular data and transcripts. We present the migration of a live, production conversational recommendation system from a gradient-boosted multiclass model to a pairwise-binary deep recommender. Because this system is critical to ecosystem growth initiatives and downstream features like dynamic pitching–surfacing the most relevant pitch text to a support
agent in real time during a live customer conversation–maintaining live recommendation quality was a non-negotiable constraint. We detail the techniques that made this migration successful–reformulating recommendation as pairwise binary prediction to learn jointly from user and item features, and enhancing learned representations via negative sampling and noise injection. To efficiently incorporate long, live conversation context, we apply
attention pooling over transcript chunks and
benchmark it against TF-IDF and sentence-
embedding baselines. Finally, we explore multiple architectures (including two-tower models, DeepFM, and their variants) and loss functions such as contrastive loss. Evaluating against a CatBoost baseline across all conversational stages, we demonstrate that our approach achieves parity at conversation beginning and outperforms at later conversational stages.
Read the full abstract ▾
All authors Sonia Sharma, Jeyendran Balakrishnan, Shreya Rajpal, Swapnil Parekh, Nagaraj Janardhana, Andrew Mattarella-Micke