Showing 1–50 of 574 results for author: Xiao, H

Adversarial Label Invariant Graph Data Augmentations for Out-of-Distribution Generalization

Authors: Simon Zhang, Ryan P. DeMilt, Kun Jin, Cathy H. Xia

Abstract: Out-of-distribution (OoD) generalization occurs when representation learning encounters a distribution shift. This occurs frequently in practice when training and testing data come from different environments. Covariate shift is a type of distribution shift that occurs only in the input data, while the concept distribution stays invariant. We propose RIA - Regularization for Invariance with Advers… ▽ More Out-of-distribution (OoD) generalization occurs when representation learning encounters a distribution shift. This occurs frequently in practice when training and testing data come from different environments. Covariate shift is a type of distribution shift that occurs only in the input data, while the concept distribution stays invariant. We propose RIA - Regularization for Invariance with Adversarial training, a new method for OoD generalization under convariate shift. Motivated by an analogy to $Q$-learning, it performs an adversarial exploration for training data environments. These new environments are induced by adversarial label invariant data augmentations that prevent a collapse to an in-distribution trained learner. It works with many existing OoD generalization methods for covariate shift that can be formulated as constrained optimization problems. We develop an alternating gradient descent-ascent algorithm to solve the problem, and perform extensive experiments on OoD graph classification for various kinds of synthetic and natural distribution shifts. We demonstrate that our method can achieve high accuracy compared with OoD baselines. △ Less

Submitted 9 April, 2026; originally announced April 2026.

Comments: 21 pages, 3 figures, accepted at ICML SCIS 2023

3DrawAgent: Teaching LLM to Draw in 3D with Early Contrastive Experience

Authors: Hongcan Xiao, Xinyue Xiao, Yilin Wang, Yue Zhang, Yonggang Qi

Abstract: Sketching in 3D space enables expressive reasoning about shape, structure, and spatial relationships, yet generating 3D sketches through natural language remains a major challenge. In this work, we introduce 3DrawAgent, a training-free, language-driven framework for 3D sketch generation that leverages large language models (LLMs) to sequentially draw 3D Bezier curves under geometric feedback. Unli… ▽ More Sketching in 3D space enables expressive reasoning about shape, structure, and spatial relationships, yet generating 3D sketches through natural language remains a major challenge. In this work, we introduce 3DrawAgent, a training-free, language-driven framework for 3D sketch generation that leverages large language models (LLMs) to sequentially draw 3D Bezier curves under geometric feedback. Unlike prior 2D sketch agents, our method introduces a relative experience optimization strategy that adapts the recently proposed Group Reward Policy Optimization (GRPO) paradigm. Instead of relying on explicit ground-truth supervision, we construct pairwise comparisons among generated sketches, with each pair consisting of a relatively better and a worse result based on CLIP-based perceptual rewards and LLM-based fine-grained qualitative assessment. These experiences are then used to iteratively refine the prior knowledge of 3D drawing, enabling black-box reinforcement of the model's 3D awareness. This design allows our model to self-improve its spatial understanding and drawing quality without parameter updates. Experiments show that 3DrawAgent can generate complex and coherent 3D Bezier sketches from diverse textual prompts, exhibit emergent geometric reasoning, and generalize to novel shapes, establishing a new paradigm for advancing the field of training-free 3D sketch intelligence. △ Less

Submitted 9 April, 2026; originally announced April 2026.

Comments: CVPR 2026 Highlight

A Complete Characterization of Convexity in Flow Games

Authors: Han Xiao, Luying Zhang, Qizhi Fang

Abstract: We investigate the convexity of cooperative games arising from network flow problems. While it is well-known that flow games are totally balanced, a complete characterization of their convexity has remained an open problem. In this paper, we provide a necessary and sufficient characterization of the networks that induce convex flow games. We show that a flow game is convex if and only if the under… ▽ More We investigate the convexity of cooperative games arising from network flow problems. While it is well-known that flow games are totally balanced, a complete characterization of their convexity has remained an open problem. In this paper, we provide a necessary and sufficient characterization of the networks that induce convex flow games. We show that a flow game is convex if and only if the underlying network is acyclic and admits an arc cover by $s$-$t$ paths that are disjoint at their bottleneck arcs. Specifically, every bottleneck arc must belong to exactly one path, and every non-bottleneck arc must possess sufficient capacity. To derive this characterization, we establish six structural properties of convex flow games. Additionally, we prove that our characterization can be verified efficiently, yielding a polynomial-time algorithm to recognize convex flow games. Since the class of flow games coincides exactly with the class of non-negative totally balanced games, as established by Kalai and Zemel (1982), our structural and algorithmic characterization applies to all such games, provided they are represented in their network form. △ Less

Submitted 6 April, 2026; originally announced April 2026.

MSC Class: 05C57; 91A12; 91A43; 91A46

Multimodal Backdoor Attack on VLMs for Autonomous Driving via Graffiti and Cross-Lingual Triggers

Authors: Jiancheng Wang, Lidan Liang, Yong Wang, Zengzhen Su, Haifeng Xia, Yuanting Yan, Wei Wang

Abstract: Visual language model (VLM) is rapidly being integrated into safety-critical systems such as autonomous driving, making it an important attack surface for potential backdoor attacks. Existing backdoor attacks mainly rely on unimodal, explicit, and easily detectable triggers, making it difficult to construct both covert and stable attack channels in autonomous driving scenarios. GLA introduces two… ▽ More Visual language model (VLM) is rapidly being integrated into safety-critical systems such as autonomous driving, making it an important attack surface for potential backdoor attacks. Existing backdoor attacks mainly rely on unimodal, explicit, and easily detectable triggers, making it difficult to construct both covert and stable attack channels in autonomous driving scenarios. GLA introduces two naturalistic triggers: graffiti-based visual patterns generated via stable diffusion inpainting, which seamlessly blend into urban scenes, and cross-language text triggers, which introduce distributional shifts while maintaining semantic consistency to build robust language-side trigger signals. Experiments on DriveVLM show that GLA requires only a 10\% poisoning ratio to achieve a 90\% Attack Success Rate (ASR) and a 0\% False Positive Rate (FPR). More insidiously, the backdoor does not weaken the model on clean tasks, but instead improves metrics such as BLEU-1, making it difficult for traditional performance-degradation-based detection methods to identify the attack. This study reveals underestimated security threats in self-driving VLMs and provides a new attack paradigm for backdoor evaluation in safety-critical multimodal systems. △ Less

Submitted 6 April, 2026; originally announced April 2026.

Comments: This is a submission to the "Pattern Analysis and Applications". The manuscript includes 14 pages and 6 figures. All authors have approved the submission, and there is no conflict of interest to declare

SCC-Loc: A Unified Semantic Cascade Consensus Framework for UAV Thermal Geo-Localization

Authors: Xiaoran Zhang, Yu Liu, Jinyu Liang, Kangqiushi Li, Zhiwei Huang, Huaxin Xiao

Abstract: Cross-modal Thermal Geo-localization (TG) provides a robust, all-weather solution for Unmanned Aerial Vehicles (UAVs) in Global Navigation Satellite System (GNSS)-denied environments. However, profound thermal-visible modality gaps introduce severe feature ambiguity, systematically corrupting conventional coarse-to-fine registration. To dismantle this bottleneck, we propose SCC-Loc, a unified Sema… ▽ More Cross-modal Thermal Geo-localization (TG) provides a robust, all-weather solution for Unmanned Aerial Vehicles (UAVs) in Global Navigation Satellite System (GNSS)-denied environments. However, profound thermal-visible modality gaps introduce severe feature ambiguity, systematically corrupting conventional coarse-to-fine registration. To dismantle this bottleneck, we propose SCC-Loc, a unified Semantic-Cascade-Consensus localization framework. By sharing a single DINOv2 backbone across global retrieval and MINIMA$_{\text{RoMa}}$ matching, it minimizes memory footprint and achieves zero-shot, highly accurate absolute position estimation. Specifically, we tackle modality ambiguity by introducing three cohesive components. First, we design the Semantic-Guided Viewport Alignment (SGVA) module to adaptively optimize satellite crop regions, effectively correcting initial spatial deviations. Second, we develop the Cascaded Spatial-Adaptive Texture-Structure Filtering (C-SATSF) mechanism to explicitly enforce geometric consistency, thereby eradicating dense cross-modal outliers. Finally, we propose the Consensus-Driven Reliability-Aware Position Selection (CD-RAPS) strategy to derive the optimal solution through a synergy of physically constrained pose optimization. To address data scarcity, we construct Thermal-UAV, a comprehensive dataset providing 11,890 diverse thermal queries referenced against a large-scale satellite ortho-photo and corresponding spatially aligned Digital Surface Model (DSM). Extensive experiments demonstrate that SCC-Loc establishes a new state-of-the-art, suppressing the mean localization error to 9.37 m and providing a 7.6-fold accuracy improvement within a strict 5-m threshold over the strongest baseline. Code and dataset are available at https://github.com/FloralHercules/SCC-Loc. △ Less

Submitted 3 April, 2026; originally announced April 2026.

Comments: 15 pages, 4 figures. Submitted to IEEE J-STARS

Architecting Secure AI Agents: Perspectives on System-Level Defenses Against Indirect Prompt Injection Attacks

Authors: Chong Xiang, Drew Zagieboylo, Shaona Ghosh, Sanjay Kariyappa, Kai Greshake, Hanshen Xiao, Chaowei Xiao, G. Edward Suh

Abstract: AI agents, predominantly powered by large language models (LLMs), are vulnerable to indirect prompt injection, in which malicious instructions embedded in untrusted data can trigger dangerous agent actions. This position paper discusses our vision for system-level defenses against indirect prompt injection attacks. We articulate three positions: (1) dynamic replanning and security policy updates a… ▽ More AI agents, predominantly powered by large language models (LLMs), are vulnerable to indirect prompt injection, in which malicious instructions embedded in untrusted data can trigger dangerous agent actions. This position paper discusses our vision for system-level defenses against indirect prompt injection attacks. We articulate three positions: (1) dynamic replanning and security policy updates are often necessary for dynamic tasks and realistic environments; (2) certain context-dependent security decisions would still require LLMs (or other learned models), but should only be made within system designs that strictly constrain what the model can observe and decide; (3) in inherently ambiguous cases, personalization and human interaction should be treated as core design considerations. In addition to our main positions, we discuss limitations of existing benchmarks that can create a false sense of utility and security. We also highlight the value of system-level defenses, which serve as the skeleton of agentic systems by structuring and controlling agent behaviors, integrating rule-based and model-based security checks, and enabling more targeted research on model robustness and human interaction. △ Less

Submitted 31 March, 2026; originally announced March 2026.

Security-Spectral Efficiency Tradeoff in STAR-RIS RSMA: A Max-Min Fairness Framework

Authors: Huiyun Xia, Yijie Mao, Sai Xu, Shuai Han, Hongbo Zhu

Abstract: Simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RISs) enable full-space coverage but also expose wireless transmissions to security from multiple spatial directions. This paper investigates a STAR-RIS-assisted secure RSMA system where both internal and external eavesdroppers may coexist in the transmission and reflection regions. In such a scenario, the RSMA co… ▽ More Simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RISs) enable full-space coverage but also expose wireless transmissions to security from multiple spatial directions. This paper investigates a STAR-RIS-assisted secure RSMA system where both internal and external eavesdroppers may coexist in the transmission and reflection regions. In such a scenario, the RSMA common stream simultaneously serves legitimate users, impairs external eavesdroppers, and avoids assisting internal eavesdroppers, leading to a challenging trade-off between spectral efficiency and confidentiality. To address this issue, we formulate a max-min fairness problem under secrecy constraints and develop an iterative algorithm to jointly optimize transmit beamforming and STAR-RIS phase shifts. Simulation results demonstrate that the proposed scheme improves spectral efficiency while maintaining confidentiality. △ Less

Submitted 31 March, 2026; v1 submitted 27 March, 2026; originally announced March 2026.

Hessian-informed machine learning interatomic potential towards bridging theory and experiments

Authors: Bangchen Yin, Jian Ouyang, Zhen Fan, Kailai Lin, Hanshi Hu, Dingshun Lv, Weiluo Ren, Hai Xiao, Ji Chen, Changsu Cao

Abstract: Local curvature of potential energy surfaces is critical for predicting certain experimental observables of molecules and materials from first principles, yet it remains far beyond reach for complex systems. In this work, we introduce a Hessian-informed Machine Learning Interatomic Potential (Hi-MLIP) that captures such curvature reliably, thereby enabling accurate analysis of associated thermodyn… ▽ More Local curvature of potential energy surfaces is critical for predicting certain experimental observables of molecules and materials from first principles, yet it remains far beyond reach for complex systems. In this work, we introduce a Hessian-informed Machine Learning Interatomic Potential (Hi-MLIP) that captures such curvature reliably, thereby enabling accurate analysis of associated thermodynamic and kinetic phenomena. To make Hessian supervision practically viable, we develop a highly efficient training protocol, termed Hessian INformed Training (HINT), achieving two to four orders of magnitude reduction for the requirement of expensive Hessian labels. HINT integrates critical techniques, including Hessian pre-training, configuration sampling, curriculum learning and stochastic projection Hessian loss. Enabled by HINT, Hi-MLIP significantly improves transition-state search and brings Gibbs free-energy predictions close to chemical accuracy especially in data-scarce regimes. Our framework also enables accurate treatment of strongly anharmonic hydrides, reproducing phonon renormalization and superconducting critical temperatures in close agreement with experiment while bypassing the computational bottleneck of anharmonic calculations. These results establish a practical route to enhancing curvature awareness of machine learning interatomic potentials, bridging simulation and experimental observables across a wide range of systems. △ Less

Submitted 26 March, 2026; originally announced March 2026.

Comments: 13 pages, 4 figures

Foveated Diffusion: Efficient Spatially Adaptive Image and Video Generation

Authors: Brian Chao, Lior Yariv, Howard Xiao, Gordon Wetzstein

Abstract: Diffusion and flow matching models have unlocked unprecedented capabilities for creative content creation, such as interactive image and streaming video generation. The growing demand for higher resolutions, frame rates, and context lengths, however, makes efficient generation increasingly challenging, as computational complexity grows quadratically with the number of generated tokens. Our work se… ▽ More Diffusion and flow matching models have unlocked unprecedented capabilities for creative content creation, such as interactive image and streaming video generation. The growing demand for higher resolutions, frame rates, and context lengths, however, makes efficient generation increasingly challenging, as computational complexity grows quadratically with the number of generated tokens. Our work seeks to optimize the efficiency of the generation process in settings where the user's gaze location is known or can be estimated, for example, by using eye tracking. In these settings, we leverage the eccentricity-dependent acuity of human vision: while a user perceives very high-resolution visual information in a small region around their gaze location (the foveal region), the ability to resolve detail quickly degrades in the periphery of the visual field. Our approach starts with a mask modeling the foveated resolution to allocate tokens non-uniformly, assigning higher token density to foveal regions and lower density to peripheral regions. An image or video is generated in a mixed-resolution token setting, yielding results perceptually indistinguishable from full-resolution generation, while drastically reducing the token count and generation time. To this end, we develop a principled mechanism for constructing mixed-resolution tokens directly from high-resolution data, allowing a foveated diffusion model to be post-trained from an existing base model while maintaining content consistency across resolutions. We validate our approach through extensive analysis and a carefully designed user study, demonstrating the efficacy of foveation as a practical and scalable axis for efficient generation. △ Less

Submitted 24 March, 2026; originally announced March 2026.

Comments: Project website at https://bchao1.github.io/foveated-diffusion

Cross-Granularity Representations for Biological Sequences: Insights from ESM and BiGCARP

Authors: Hanlin Xiao, Rainer Breitling, Eriko Takano, Mauricio A. Álvarez

Abstract: Recent advances in general-purpose foundation models have stimulated the development of large biological sequence models. While natural language shows symbolic granularity (characters, words, sentences), biological sequences exhibit hierarchical granularity whose levels (nucleotides, amino acids, protein domains, genes) further encode biologically functional information. In this paper, we investig… ▽ More Recent advances in general-purpose foundation models have stimulated the development of large biological sequence models. While natural language shows symbolic granularity (characters, words, sentences), biological sequences exhibit hierarchical granularity whose levels (nucleotides, amino acids, protein domains, genes) further encode biologically functional information. In this paper, we investigate the integration of cross-granularity knowledge from models through a case study of BiGCARP, a Pfam domain-level model for biosynthetic gene clusters, and ESM, an amino acid-level protein language model. Using representation analysis tools and a set of probe tasks, we first explain why a straightforward cross-model embedding initialization fails to improve downstream performance in BiGCARP, and show that deeper-layer embeddings capture a more contextual and faithful representation of the model's learned knowledge. Furthermore, we demonstrate that representations at different granularities encode complementary biological knowledge, and that combining them yields measurable performance gains in intermediate-level prediction tasks. Our findings highlight cross-granularity integration as a promising strategy for improving both the performance and interpretability of biological foundation models. △ Less

Submitted 21 March, 2026; originally announced March 2026.

Comments: 9 pages, 4 figures, published in 2025 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

Journal ref: Proc. IEEE BIBM (2025) 6936-6943

S3T-Former: A Purely Spike-Driven State-Space Topology Transformer for Skeleton Action Recognition

Authors: Naichuan Zheng, Hailun Xia, Zepeng Sun, Weiyi Li, Yujia Wang

Abstract: Skeleton-based action recognition is crucial for multimedia applications but heavily relies on power-hungry Artificial Neural Networks (ANNs), limiting their deployment on resource-constrained edge devices. Spiking Neural Networks (SNNs) provide an energy-efficient alternative; however, existing spiking models for skeleton data often compromise the intrinsic sparsity of SNNs by resorting to dense… ▽ More Skeleton-based action recognition is crucial for multimedia applications but heavily relies on power-hungry Artificial Neural Networks (ANNs), limiting their deployment on resource-constrained edge devices. Spiking Neural Networks (SNNs) provide an energy-efficient alternative; however, existing spiking models for skeleton data often compromise the intrinsic sparsity of SNNs by resorting to dense matrix aggregations, heavy multimodal fusion modules, or non-sparse frequency domain transformations. Furthermore, they severely suffer from the short-term amnesia of spiking neurons. In this paper, we propose the Spiking State-Space Topology Transformer (S3T-Former), which, to the best of our knowledge, is the first purely spike-driven Transformer architecture specifically designed for energy-efficient skeleton action recognition. Rather than relying on heavy fusion overhead, we formulate a Multi-Stream Anatomical Spiking Embedding (M-ASE) that acts as a generalized kinematic differential operator, elegantly transforming multimodal skeleton features into heterogeneous, highly sparse event streams. To achieve true topological and temporal sparsity, we introduce Lateral Spiking Topology Routing (LSTR) for on-demand conditional spike propagation, and a Spiking State-Space (S3) Engine to systematically capture long-range temporal dynamics without non-sparse spectral workarounds. Extensive experiments on multiple large-scale datasets demonstrate that S3T-Former achieves highly competitive accuracy while theoretically reducing energy consumption compared to classic ANNs, establishing a new state-of-the-art for energy-efficient neuromorphic action recognition. △ Less

Submitted 19 March, 2026; v1 submitted 17 March, 2026; originally announced March 2026.

Beyond Outliers: A Data-Free Layer-wise Mixed-Precision Quantization Approach Driven by Numerical and Structural Dual-Sensitivity

Authors: Hengyuan Zhang, Xinrong Chen, Zunhai Su, Xiao Liang, Jing Xiong, Wendong Xu, He Xiao, Chaofan Tao, Wei Zhang, Ruobing Xie, Lei Jiang, Hayden Kwok-Hay So, Ngai Wong

Abstract: Layer-wise mixed-precision quantization (LMPQ) enables effective compression under extreme low-bit settings by allocating higher precision to sensitive layers. However, existing methods typically treat all intra-layer weight modules uniformly and rely on a single numerical property when estimating sensitivity, overlooking their distinct operational roles and structural characteristics. To address… ▽ More Layer-wise mixed-precision quantization (LMPQ) enables effective compression under extreme low-bit settings by allocating higher precision to sensitive layers. However, existing methods typically treat all intra-layer weight modules uniformly and rely on a single numerical property when estimating sensitivity, overlooking their distinct operational roles and structural characteristics. To address this, we propose NSDS, a novel calibration-free LMPQ framework driven by Numerical and Structural Dual-Sensitivity. Specifically, it first mechanistically decomposes each layer into distinct operational roles and quantifies their sensitivity from both numerical and structural perspectives. These dual-aspect scores are then aggregated into a unified layer-wise metric through a robust aggregation scheme based on MAD-Sigmoid and Soft-OR to guide bit allocation. Extensive experiments demonstrate that NSDS consistently achieves superior performance compared to various baselines across diverse models and downstream tasks, without relying on any calibration data. △ Less

Submitted 18 March, 2026; originally announced March 2026.

Evo-Retriever: LLM-Guided Curriculum Evolution with Viewpoint-Pathway Collaboration for Multimodal Document Retrieval

Authors: Weiqing Li, Jinyue Guo, Yaqi Wang, Haiyang Xiao, Yuewei Zhang, Guohua Liu, Hao Henry Wang

Abstract: Visual-language models (VLMs) excel at data mappings, but real-world document heterogeneity and unstructuredness disrupt the consistency of cross-modal embeddings. Recent late-interaction methods enhance image-text alignment through multi-vector representations, yet traditional training with limited samples and static strategies cannot adapt to the model's dynamic evolution, causing cross-modal re… ▽ More Visual-language models (VLMs) excel at data mappings, but real-world document heterogeneity and unstructuredness disrupt the consistency of cross-modal embeddings. Recent late-interaction methods enhance image-text alignment through multi-vector representations, yet traditional training with limited samples and static strategies cannot adapt to the model's dynamic evolution, causing cross-modal retrieval confusion. To overcome this, we introduce Evo-Retriever, a retrieval framework featuring an LLM-guided curriculum evolution built upon a novel Viewpoint-Pathway collaboration. First, we employ multi-view image alignment to enhance fine-grained matching via multi-scale and multi-directional perspectives. Then, a bidirectional contrastive learning strategy generates "hard queries" and establishes complementary learning paths for visual and textual disambiguation to rebalance supervision. Finally, the model-state summary from the above collaboration is fed into an LLM meta-controller, which adaptively adjusts the training curriculum using expert knowledge to promote the model's evolution. On ViDoRe V2 and MMEB (VisDoc), Evo-Retriever achieves state-of-the-art performance, with nDCG@5 scores of 65.2% and 77.1%. △ Less

Submitted 17 March, 2026; originally announced March 2026.

Comments: Accepted by CVPR2026

MBD: A Model-Based Debiasing Framework Across User, Content, and Model Dimensions

Authors: Yuantong Li, Lei Yuan, Zhihao Zheng, Weimiao Wu, Songbin Liu, Jeong Min Lee, Ali Selman Aydin, Shaofeng Deng, Junbo Chen, Xinyi Zhang, Hongjing Xia, Sam Fieldman, Matthew Kosko, Wei Fu, Du Zhang, Peiyu Yang, Albert Jin Chung, Xianlei Qiu, Miao Yu, Zhongwei Teng, Hao Chen, Sunny Baek, Hui Tang, Yang Lv, Renze Wang , et al. (5 additional authors not shown)

Abstract: Modern recommendation systems rank candidates by aggregating multiple behavioral signals through a value model. However, many commonly used signals are inherently affected by heterogeneous biases. For example, watch time naturally favors long-form content, loop rate favors short - form content, and comment probability favors videos over images. Such biases introduce two critical issues: (1) value… ▽ More Modern recommendation systems rank candidates by aggregating multiple behavioral signals through a value model. However, many commonly used signals are inherently affected by heterogeneous biases. For example, watch time naturally favors long-form content, loop rate favors short - form content, and comment probability favors videos over images. Such biases introduce two critical issues: (1) value model scores may be systematically misaligned with users' relative preferences - for instance, a seemingly low absolute like probability may represent exceptionally strong interest for a user who rarely engages; and (2) changes in value modeling rules can trigger abrupt and undesirable ecosystem shifts. In this work, we ask a fundamental question: can biased behavioral signals be systematically transformed into unbiased signals, under a user - defined notion of ``unbiasedness'', that are both personalized and adaptive? We propose a general, model-based debiasing (MBD) framework that addresses this challenge by augmenting it with distributional modeling. By conditioning on a flexible subset of features (partial feature set), we explicitly estimate the contextual mean and variance of the engagement distribution for arbitrary cohorts (e.g., specific video lengths or user regions) directly alongside the main prediction. This integration allows the framework to convert biased raw signals into unbiased representations, enabling the construction of higher-level, calibrated signals (such as percentiles or z - scores) suitable for the value model. Importantly, the definition of unbiasedness is flexible and controllable, allowing the system to adapt to different personalization objectives and modeling preferences. Crucially, this is implemented as a lightweight, built-in branch of the existing MTML ranking model, requiring no separate serving infrastructure. △ Less

Submitted 15 March, 2026; originally announced March 2026.

Generative Inverse Design of Cold Metals for Low-Power Electronics

Authors: Kedeng Wu, Yucheng Zhu, Yan Chen, Bizhu Zhang, Shuyu Liu, Xiaobin Deng, Yabei Wu, Liangliang Zhu, Hang Xiao

Abstract: Cold metals are a class of metals with an intrinsic energy gap located close to the Fermi level, which enables cold-carrier injection for steep-slope transistors and is therefore promising for low-power electronic applications. High-throughput screening has revealed 252 three-dimensional (3D) cold metals in the Materials Project database, but database searches are inherently limited to known compo… ▽ More Cold metals are a class of metals with an intrinsic energy gap located close to the Fermi level, which enables cold-carrier injection for steep-slope transistors and is therefore promising for low-power electronic applications. High-throughput screening has revealed 252 three-dimensional (3D) cold metals in the Materials Project database, but database searches are inherently limited to known compounds. Here we present an inverse-design workflow that generates 3D cold metals using MatterGPT, a conditional autoregressive Transformer trained on SLICES, an invertible and symmetry-invariant crystal string representation. We curate a training set of 26,309 metallic structures labeled with energy above hull and a unified band-edge distance descriptor that merges p-type and n-type cold-metal characteristics to address severe label imbalance. Property-conditioned generation targeting thermodynamic stability and 50-500 meV band-edge distances produces 148,506 unique candidates; 92.1% are successfully reconstructed to 3D structures and down-selected by symmetry, uniqueness and novelty filters, followed by high-throughput DFT validation. We identify 257 cold metals verified as novel with respect to the Materials Project database, with gaps around the Fermi level spanning 50-500 meV. First-principles phonon, electronic-structure, and work-function calculations for representative candidates confirm dynamical stability and contact-relevant work functions. Our results demonstrate that SLICES-enabled generative transformers can expand the chemical space of cold metals beyond high-throughput screening, providing a route to low-power electronic materials discovery. △ Less

Submitted 14 March, 2026; originally announced March 2026.

CogSearch: A Cognitive-Aligned Multi-Agent Framework for Proactive Decision Support in E-Commerce Search

Authors: Zhouwei Zhai, Mengxiang Chen, Haoyun Xia, Jin Li, Renquan Zhou, Min Yang

Abstract: Modern e-commerce search engines, largely rooted in passive retrieval-and-ranking models, frequently fail to support complex decision-making, leaving users overwhelmed by cognitive friction. In this paper, we introduce CogSearch, a novel cognitive-oriented multi-agent framework that reimagines e-commerce search as a proactive decision support system. By synergizing four specialized agents, CogSear… ▽ More Modern e-commerce search engines, largely rooted in passive retrieval-and-ranking models, frequently fail to support complex decision-making, leaving users overwhelmed by cognitive friction. In this paper, we introduce CogSearch, a novel cognitive-oriented multi-agent framework that reimagines e-commerce search as a proactive decision support system. By synergizing four specialized agents, CogSearch mimics human cognitive workflows: it decomposes intricate user intents, fuses heterogeneous knowledge across internal and external sources, and delivers highly actionable insights. Our offline benchmarks validate CogSearch's excellence in consultative and complex search scenarios. Extensive online A/B testing on JD.com demonstrates the system's transformative impact: it reduced decision costs by 5% and achieved a 0.41% increase in overall UCVR, with a remarkable 30% surge in conversion for decision-heavy queries. CogSearch represents a fundamental shift in information retrieval, moving beyond traditional relevance-centric paradigms toward a future of holistic, collaborative decision intelligence. △ Less

Submitted 12 March, 2026; originally announced March 2026.

Disentangling Similarity and Relatedness in Topic Models

Authors: Hanlin Xiao, Mauricio A. Álvarez, Rainer Breitling

Abstract: The recent advancement of large language models has spurred a growing trend of integrating pre-trained language model (PLM) embeddings into topic models, fundamentally reshaping how topics capture semantic structure. Classical models such as Latent Dirichlet Allocation (LDA) derive topics from word co-occurrence statistics, whereas PLM-augmented models anchor these statistics to pre-trained embedd… ▽ More The recent advancement of large language models has spurred a growing trend of integrating pre-trained language model (PLM) embeddings into topic models, fundamentally reshaping how topics capture semantic structure. Classical models such as Latent Dirichlet Allocation (LDA) derive topics from word co-occurrence statistics, whereas PLM-augmented models anchor these statistics to pre-trained embedding spaces, imposing a prior that also favours clustering of semantically similar words. This structural difference can be captured by the psycholinguistic dimensions of thematic relatedness and taxonomic similarity of the topic words. To disentangle these dimensions in topic models, we construct a large synthetic benchmark of word pairs using LLM-based annotation to train a neural scoring function. We apply this scorer to a comprehensive evaluation across multiple corpora and topic model families, revealing that different model families capture distinct semantic structure in their topics. We further demonstrate that similarity and relatedness scores successfully predict downstream task performance depending on task requirements. This paper establishes similarity and relatedness as essential axes for topic model evaluation and provides a reliable pipeline for characterising these across model families and corpora. △ Less

Submitted 11 March, 2026; originally announced March 2026.

Comments: 22 pages, 6 figures, 14 tables

Agentic Neurosymbolic Collaboration for Mathematical Discovery: A Case Study in Combinatorial Design

Authors: Hai Xia, Carla P. Gomes, Bart Selman, Stefan Szeider

Abstract: We study mathematical discovery through the lens of neurosymbolic reasoning, where an AI agent powered by a large language model (LLM), coupled with symbolic computation tools, and human strategic direction, jointly produced a new result in combinatorial design theory. The main result of this human-AI collaboration is a tight lower bound on the imbalance of Latin squares for the notoriously diffic… ▽ More We study mathematical discovery through the lens of neurosymbolic reasoning, where an AI agent powered by a large language model (LLM), coupled with symbolic computation tools, and human strategic direction, jointly produced a new result in combinatorial design theory. The main result of this human-AI collaboration is a tight lower bound on the imbalance of Latin squares for the notoriously difficult case $n \equiv 1 \pmod{3}$. We reconstruct the discovery process from detailed interaction logs spanning multiple sessions over several days and identify the distinct cognitive contributions of each component. The AI agent proved effective at uncovering hidden structure and generating hypotheses. The symbolic component consists of computer algebra, constraint solvers, and simulated annealing, which provides rigorous verification and exhaustive enumeration. Human steering supplied the critical research pivot that transformed a dead end into a productive inquiry. Our analysis reveals that multi-model deliberation among frontier LLMs proved reliable for criticism and error detection but unreliable for constructive claims. The resulting human-AI mathematical contribution, a tight lower bound of $4n(n{-}1)/9$, is achieved via a novel class of near-perfect permutations. The bound was formally verified in Lean 4. Our experiments show that neurosymbolic systems can indeed produce genuine discoveries in pure mathematics. △ Less

Submitted 9 March, 2026; originally announced March 2026.

PIRA-Bench: A Transition from Reactive GUI Agents to GUI-based Proactive Intent Recommendation Agents

Authors: Yuxiang Chai, Shunye Tang, Han Xiao, Rui Liu, Hongsheng Li

Abstract: Current Graphical User Interface (GUI) agents operate primarily under a reactive paradigm: a user must provide an explicit instruction for the agent to execute a task. However, an intelligent AI assistant should be proactive, which is capable of anticipating user intentions directly from continuous visual inputs, such as mobile or desktop screenshots, and offering timely recommendations without ex… ▽ More Current Graphical User Interface (GUI) agents operate primarily under a reactive paradigm: a user must provide an explicit instruction for the agent to execute a task. However, an intelligent AI assistant should be proactive, which is capable of anticipating user intentions directly from continuous visual inputs, such as mobile or desktop screenshots, and offering timely recommendations without explicit user prompting. Transitioning to this proactive paradigm presents significant challenges. Real-world screen activity is rarely linear; it consists of long-horizon trajectories fraught with noisy browsing, meaningless actions, and multithreaded task-switching. To address this gap, we introduce PIRA-Bench (Proactive Intent Recommendation Agent Benchmark), a novel benchmark for evaluating multimodal large language models (MLLMs) on continuous, weakly-supervised visual inputs. Unlike reactive datasets, PIRA-Bench features complex trajectories with multiple interleaved intents and noisy segments with various user profile contexts, challenging agents to detect actionable events while fitting to user preferences. Furthermore, we propose the PIRF baseline, a memory-aware, state-tracking framework that empowers general MLLMs to manage multiple task threads and handle misleading visual inputs. PIRA-Bench serves as an initial step toward robust and proactive GUI-based personal assistants. △ Less

Submitted 9 March, 2026; originally announced March 2026.

FuzzingRL: Reinforcement Fuzz-Testing for Revealing VLM Failures

Authors: Jiajun Xu, Jiageng Mao, Ang Qi, Weiduo Yuan, Alexander Romanus, Helen Xia, Vitor Campagnolo Guizilini, Yue Wang

Abstract: Vision Language Models (VLMs) are prone to errors, and identifying where these errors occur is critical for ensuring the reliability and safety of AI systems. In this paper, we propose an approach that automatically generates questions designed to deliberately induce incorrect responses from VLMs, thereby revealing their vulnerabilities. The core of this approach lies in fuzz testing and reinforce… ▽ More Vision Language Models (VLMs) are prone to errors, and identifying where these errors occur is critical for ensuring the reliability and safety of AI systems. In this paper, we propose an approach that automatically generates questions designed to deliberately induce incorrect responses from VLMs, thereby revealing their vulnerabilities. The core of this approach lies in fuzz testing and reinforcement finetuning: we transform a single input query into a large set of diverse variants through vision and language fuzzing. Based on the fuzzing outcomes, the question generator is further instructed by adversarial reinforcement fine-tuning to produce increasingly challenging queries that trigger model failures. With this approach, we can consistently drive down a target VLM's answer accuracy -- for example, the accuracy of Qwen2.5-VL-32B on our generated questions drops from 86.58\% to 65.53\% in four RL iterations. Moreover, a fuzzing policy trained against a single target VLM transfers to multiple other VLMs, producing challenging queries that degrade their performance as well. △ Less

Submitted 17 February, 2026; originally announced March 2026.

Comments: 18 pages, 4 figures. † These authors jointly supervised this work: Jiageng Mao and Yue Wang

VizCrit: Exploring Strategies for Displaying Computational Feedback in a Visual Design Tool

Authors: Mingyi Li, Mengyi Chen, Sarah Luo, Yining Cao, Haijun Xia, Maitraye Das, Steven P. Dow, Jane L. E

Abstract: Visual design instructors often provide multi-modal feedback, mixing annotations with text. Prior theory emphasizes the importance of actionable feedback, where "actionability" lies on a spectrum--from surfacing relevant design concepts to suggesting concrete fixes. How might creativity tools implement annotations that support such feedback, and how does the actionability of feedback impact novice… ▽ More Visual design instructors often provide multi-modal feedback, mixing annotations with text. Prior theory emphasizes the importance of actionable feedback, where "actionability" lies on a spectrum--from surfacing relevant design concepts to suggesting concrete fixes. How might creativity tools implement annotations that support such feedback, and how does the actionability of feedback impact novices' process-related behaviors, perceptions of creativity, learning of design principles, and overall outcomes? We introduce VizCrit, a system for providing computational feedback that supports the actionability spectrum, realized through algorithmic issue detection and visual annotation generation. In a between-subjects study (N=36), novices revised a design under one of three conditions: textbook-based, awareness-centered, or solution-centered feedback. We found that solution-centered feedback led to fewer design issues and higher self-perceived creativity compared with textbook-based feedback, although expert ratings on creativity showed no significant differences. We discuss the implications for AI in Creativity Support Tools, including the potential of calibrating feedback actionability to help novices balance productivity with learning, growth, and developing design awareness. △ Less

Submitted 4 March, 2026; originally announced March 2026.

mlx-vis: GPU-Accelerated Dimensionality Reduction and Visualization on Apple Silicon

Authors: Han Xiao

Abstract: mlx-vis implements eight dimensionality reduction methods -- UMAP, t-SNE, PaCMAP, LocalMAP, TriMap, DREAMS, CNE, MMAE -- and NNDescent k-NN graph construction entirely in MLX for Apple Silicon Metal GPU. A built-in GPU renderer produces scatter plots and smooth animations via hardware H.264 encoding. On Fashion-MNIST (70K points, M3 Ultra), seven of eight methods embed in 2.0-4.7s and render 800-f… ▽ More mlx-vis implements eight dimensionality reduction methods -- UMAP, t-SNE, PaCMAP, LocalMAP, TriMap, DREAMS, CNE, MMAE -- and NNDescent k-NN graph construction entirely in MLX for Apple Silicon Metal GPU. A built-in GPU renderer produces scatter plots and smooth animations via hardware H.264 encoding. On Fashion-MNIST (70K points, M3 Ultra), seven of eight methods embed in 2.0-4.7s and render 800-frame animations in 1.4s. The library depends only on MLX and NumPy and is available at https://github.com/hanxiao/mlx-vis. △ Less

Submitted 20 March, 2026; v1 submitted 4 March, 2026; originally announced March 2026.

Comments: 8 pages, 8 figures. Software: https://github.com/hanxiao/mlx-vis. v3: VRAM optimization, updated benchmarks, added LocalMAP and MMAE methods

Towards Privacy-Preserving LLM Inference via Covariant Obfuscation (Technical Report)

Authors: Yu Lin, Qizhi Zhang, Wenqiang Ruan, Daode Zhang, Jue Hong, Ye Wu, Hanning Xia, Yunlong Mao, Sheng Zhong

Abstract: The rapid development of large language models (LLMs) has driven the widespread adoption of cloud-based LLM inference services, while also bringing prominent privacy risks associated with the transmission and processing of private data in remote inference. For privacy-preserving LLM inference technologies to be practically applied in industrial scenarios, three core requirements must be satisfied… ▽ More The rapid development of large language models (LLMs) has driven the widespread adoption of cloud-based LLM inference services, while also bringing prominent privacy risks associated with the transmission and processing of private data in remote inference. For privacy-preserving LLM inference technologies to be practically applied in industrial scenarios, three core requirements must be satisfied simultaneously: (1) Accuracy and efficiency losses should be minimized to mitigate degradation in service experience. (2) The inference process can be run on large-scale clusters consist of heterogeneous legacy xPUs. (3) Compatibility with existing LLM infrastructures should be ensured to reuse their engineering optimizations. To the best of our knowledge, none of the existing privacy-preserving LLM inference methods satisfy all the above constraints while delivering meaningful privacy guarantees. In this paper, we propose AloePri, the first privacy-preserving LLM inference method for industrial applications. AloePri protects both the input and output data by covariant obfuscation, which jointly transforms data and model parameters to achieve better accuracy and privacy. We carefully design the transformation for each model component to ensure inference accuracy and data privacy while keeping full compatibility with existing infrastructures of Language Model as a Service. AloePri has been integrated into an industrial system for the evaluation of mainstream LLMs. The evaluation on Deepseek-V3.1-Terminus model (671B parameters) demonstrates that AloePri causes accuracy loss of 0.0%~3.5% and exhibits efficiency equivalent to that of plaintext inference. Meanwhile, AloePri successfully resists state-of-the-art attacks, with less than 5\% of tokens recovered. To the best of our knowledge, AloePri is the first method to exhibit practical applicability to large-scale models in real-world systems. △ Less

Submitted 30 March, 2026; v1 submitted 2 March, 2026; originally announced March 2026.

CMI-RewardBench: Evaluating Music Reward Models with Compositional Multimodal Instruction

Authors: Yinghao Ma, Haiwen Xia, Hewei Gao, Weixiong Chen, Yuxin Ye, Yuchen Yang, Sungkyun Chang, Mingshuo Ding, Yizhi Li, Ruibin Yuan, Simon Dixon, Emmanouil Benetos

Abstract: While music generation models have evolved to handle complex multimodal inputs mixing text, lyrics, and reference audio, evaluation mechanisms have lagged behind. In this paper, we bridge this critical gap by establishing a comprehensive ecosystem for music reward modeling under Compositional Multimodal Instruction (CMI), where the generated music may be conditioned on text descriptions, lyrics, a… ▽ More While music generation models have evolved to handle complex multimodal inputs mixing text, lyrics, and reference audio, evaluation mechanisms have lagged behind. In this paper, we bridge this critical gap by establishing a comprehensive ecosystem for music reward modeling under Compositional Multimodal Instruction (CMI), where the generated music may be conditioned on text descriptions, lyrics, and audio prompts. We first introduce CMI-Pref-Pseudo, a large-scale preference dataset comprising 110k pseudo-labeled samples, and CMI-Pref, a high-quality, human-annotated corpus tailored for fine-grained alignment tasks. To unify the evaluation landscape, we propose CMI-RewardBench, a unified benchmark that evaluates music reward models on heterogeneous samples across musicality, text-music alignment, and compositional instruction alignment. Leveraging these resources, we develop CMI reward models (CMI-RMs), a parameter-efficient reward model family capable of processing heterogeneous inputs. We evaluate their correlation with human judgments scores on musicality and alignment on CMI-Pref along with previous datasets. Further experiments demonstrate that CMI-RM not only correlates strongly with human judgments, but also enables effective inference-time scaling via top-k filtering. The necessary training data, benchmarks, and reward models are publicly available. △ Less

Submitted 4 March, 2026; v1 submitted 28 February, 2026; originally announced March 2026.

SWE-Hub: A Unified Production System for Scalable, Executable Software Engineering Tasks

Authors: Yucheng Zeng, Shupeng Li, Daxiang Dong, Ruijie Xu, Zimo Chen, Liwei Zheng, Yuxuan Li, Zhe Zhou, Haotian Zhao, Lun Tian, Heng Xiao, Tianshu Zhu, Longkun Hao, Jianmin Wu

Abstract: Progress in software-engineering agents is increasingly constrained by the scarcity of executable, scalable, and realistic data for training and evaluation. This scarcity stems from three fundamental challenges in existing pipelines: environments are brittle and difficult to reproduce across languages; synthesizing realistic, system-level bugs at scale is computationally expensive; and existing da… ▽ More Progress in software-engineering agents is increasingly constrained by the scarcity of executable, scalable, and realistic data for training and evaluation. This scarcity stems from three fundamental challenges in existing pipelines: environments are brittle and difficult to reproduce across languages; synthesizing realistic, system-level bugs at scale is computationally expensive; and existing data predominantly consists of short-horizon repairs, failing to capture long-horizon competencies like architectural consistency. We introduce \textbf{SWE-Hub}, an end-to-end system that operationalizes the data factory abstraction by unifying environment automation, scalable synthesis, and diverse task generation into a coherent production stack. At its foundation, the \textbf{Env Agent} establishes a shared execution substrate by automatically converting raw repository snapshots into reproducible, multi-language container environments with standardized interfaces. Built upon this substrate, \textbf{SWE-Scale} engine addresses the need for high-throughput generation, combining cross-language code analysis with cluster-scale validation to synthesize massive volumes of localized bug-fix instances. \textbf{Bug Agent} generates high-fidelity repair tasks by synthesizing system-level regressions involving cross-module dependencies, paired with user-like issue reports that describe observable symptoms rather than root causes. Finally, \textbf{SWE-Architect} expands the task scope from repair to creation by translating natural-language requirements into repository-scale build-a-repo tasks. By integrating these components, SWE-Hub establishes a unified production pipeline capable of continuously delivering executable tasks across the entire software engineering lifecycle. △ Less

Submitted 28 February, 2026; originally announced March 2026.

Tidynote: Always-Clear Notebook Authoring

Authors: Ruanqianqian Huang, Brian Hempel, Yining Cao, James D. Hollan, Haijun Xia, Sorin Lerner

Abstract: Recent work identified clarity as one of the top quality attributes that notebook users value, but notebooks lack support for maintaining clarity throughout the exploratory phases of the notebook authoring workflow. We propose always-clear notebook authoring that supports both clarity and exploration, and present a Jupyter implementation called Tidynote. The key to Tidynote is three-fold: (1) a sc… ▽ More Recent work identified clarity as one of the top quality attributes that notebook users value, but notebooks lack support for maintaining clarity throughout the exploratory phases of the notebook authoring workflow. We propose always-clear notebook authoring that supports both clarity and exploration, and present a Jupyter implementation called Tidynote. The key to Tidynote is three-fold: (1) a scratchpad sidebar to facilitate exploration, (2) cells movable between the notebook and the scratchpad to maintain organization, and (3) linear execution with state forks to clarify program state. An exploratory study (N=13) of open-ended data analysis tasks shows that Tidynote features holistically promote clarity throughout a notebook's lifecycle, support realistic notebook tasks, and enable novel strategies for notebook clarity. These results suggest that Tidynote supports maintaining clarity throughout the entirety of notebook authoring. △ Less

Submitted 26 February, 2026; originally announced February 2026.

Comments: Accepted at CHI 2026

SceneTransporter: Optimal Transport-Guided Compositional Latent Diffusion for Single-Image Structured 3D Scene Generation

Authors: Ling Wang, Hao-Xiang Guo, Xinzhou Wang, Fuchun Sun, Kai Sun, Pengkun Liu, Hang Xiao, Zhong Wang, Guangyuan Fu, Eric Li, Yang Liu, Yikai Wang

Abstract: We introduce SceneTransporter, an end-to-end framework for structured 3D scene generation from a single image. While existing methods generate part-level 3D objects, they often fail to organize these parts into distinct instances in open-world scenes. Through a debiased clustering probe, we reveal a critical insight: this failure stems from the lack of structural constraints within the model's int… ▽ More We introduce SceneTransporter, an end-to-end framework for structured 3D scene generation from a single image. While existing methods generate part-level 3D objects, they often fail to organize these parts into distinct instances in open-world scenes. Through a debiased clustering probe, we reveal a critical insight: this failure stems from the lack of structural constraints within the model's internal assignment mechanism. Based on this finding, we reframe the task of structured 3D scene generation as a global correlation assignment problem. To solve this, SceneTransporter formulates and solves an entropic Optimal Transport (OT) objective within the denoising loop of the compositional DiT model. This formulation imposes two powerful structural constraints. First, the resulting transport plan gates cross-attention to enforce an exclusive, one-to-one routing of image patches to part-level 3D latents, preventing entanglement. Second, the competitive nature of the transport encourages the grouping of similar patches, a process that is further regularized by an edge-based cost, to form coherent objects and prevent fragmentation. Extensive experiments show that SceneTransporter outperforms existing methods on open-world scene generation, significantly improving instance-level coherence and geometric fidelity. Code and models will be publicly available at https://2019epwl.github.io/SceneTransporter/. △ Less

Submitted 26 February, 2026; originally announced February 2026.

Comments: published at iclr 2026

Autoregressive Visual Decoding from EEG Signals

Authors: Sicheng Dai, Hongwang Xiao, Shan Yu, Qiwei Ye

Abstract: Electroencephalogram (EEG) signals have become a popular medium for decoding visual information due to their cost-effectiveness and high temporal resolution. However, current approaches face significant challenges in bridging the modality gap between EEG and image data. These methods typically rely on complex adaptation processes involving multiple stages, making it hard to maintain consistency an… ▽ More Electroencephalogram (EEG) signals have become a popular medium for decoding visual information due to their cost-effectiveness and high temporal resolution. However, current approaches face significant challenges in bridging the modality gap between EEG and image data. These methods typically rely on complex adaptation processes involving multiple stages, making it hard to maintain consistency and manage compounding errors. Furthermore, the computational overhead imposed by large-scale diffusion models limit their practicality in real-world brain-computer interface (BCI) applications. In this work, we present AVDE, a lightweight and efficient framework for visual decoding from EEG signals. First, we leverage LaBraM, a pre-trained EEG model, and fine-tune it via contrastive learning to align EEG and image representations. Second, we adopt an autoregressive generative framework based on a "next-scale prediction" strategy: images are encoded into multi-scale token maps using a pre-trained VQ-VAE, and a transformer is trained to autoregressively predict finer-scale tokens starting from EEG embeddings as the coarsest representation. This design enables coherent generation while preserving a direct connection between the input EEG signals and the reconstructed images. Experiments on two datasets show that AVDE outperforms previous state-of-the-art methods in both image retrieval and reconstruction tasks, while using only 10% of the parameters. In addition, visualization of intermediate outputs shows that the generative process of AVDE reflects the hierarchical nature of human visual perception. These results highlight the potential of autoregressive models as efficient and interpretable tools for practical BCI applications. △ Less

Submitted 9 March, 2026; v1 submitted 25 February, 2026; originally announced February 2026.

Journal ref: ICLR 2026

DHP: Efficient Scaling of MLLM Training with Dynamic Hybrid Parallelism

Authors: Yifan Niu, Han Xiao, Dongyi Liu, Wei Zhou, Jia Li

Abstract: Scaling long-context capabilities is crucial for Multimodal Large Language Models (MLLMs). However, real-world multimodal datasets are extremely heterogeneous. Existing training frameworks predominantly rely on static parallelism strategies, which suffer from severe load imbalance, redundant communication, and suboptimal hardware utilization under data heterogeneity. In this work, we propose Dynam… ▽ More Scaling long-context capabilities is crucial for Multimodal Large Language Models (MLLMs). However, real-world multimodal datasets are extremely heterogeneous. Existing training frameworks predominantly rely on static parallelism strategies, which suffer from severe load imbalance, redundant communication, and suboptimal hardware utilization under data heterogeneity. In this work, we propose Dynamic Hybrid Parallelism (DHP), an efficient parallelism strategy that adaptively reconfigures communication groups and parallelism degrees during MLLM training. We generalize the non-power-of-two parallelism degrees and develop a polynomial-time algorithm to generate near-optimal parallelism strategies with only millisecond-level overhead per training batch. DHP is able to maintain high hardware efficiency even under extreme data variability. Experimental results demonstrate that DHP significantly outperforms Megatron-LM and DeepSpeed, achieving up to 1.36 $\times$ speedup in training throughput while maintaining near-linear scaling efficiency across large-scale NPU clusters. △ Less

Submitted 25 February, 2026; originally announced February 2026.

jina-embeddings-v5-text: Task-Targeted Embedding Distillation

Authors: Mohammad Kalim Akram, Saba Sturua, Nastia Havriushenko, Quentin Herreros, Michael Günther, Maximilian Werk, Han Xiao

Abstract: Text embedding models are widely used for semantic similarity tasks, including information retrieval, clustering, and classification. General-purpose models are typically trained with single- or multi-stage processes using contrastive loss functions. We introduce a novel training regimen that combines model distillation techniques with task-specific contrastive loss to produce compact, high-perfor… ▽ More Text embedding models are widely used for semantic similarity tasks, including information retrieval, clustering, and classification. General-purpose models are typically trained with single- or multi-stage processes using contrastive loss functions. We introduce a novel training regimen that combines model distillation techniques with task-specific contrastive loss to produce compact, high-performance embedding models. Our findings suggest that this approach is more effective for training small models than purely contrastive or distillation-based training paradigms alone. Benchmark scores for the resulting models, jina-embeddings-v5-text-small and jina-embeddings-v5-text-nano, exceed or match the state-of-the-art for models of similar size. jina-embeddings-v5-text models additionally support long texts (up to 32k tokens) in many languages, and generate embeddings that remain robust under truncation and binary quantization. Model weights are publicly available, hopefully inspiring further advances in embedding model development. △ Less

Submitted 17 February, 2026; originally announced February 2026.

Comments: 14 pages, 8 figures. Model weights: https://huggingface.co/collections/jinaai/jina-embeddings-v5-text

RGA-Net: A Vision Enhancement Framework for Robotic Surgical Systems Using Reciprocal Attention Mechanisms

Authors: Quanjun Li, Weixuan Li, Han Xia, Junhua Zhou, Chi-Man Pun, Xuhang Chen

Abstract: Robotic surgical systems rely heavily on high-quality visual feedback for precise teleoperation; yet, surgical smoke from energy-based devices significantly degrades endoscopic video feeds, compromising the human-robot interface and surgical outcomes. This paper presents RGA-Net (Reciprocal Gating and Attention-fusion Network), a novel deep learning framework specifically designed for smoke remova… ▽ More Robotic surgical systems rely heavily on high-quality visual feedback for precise teleoperation; yet, surgical smoke from energy-based devices significantly degrades endoscopic video feeds, compromising the human-robot interface and surgical outcomes. This paper presents RGA-Net (Reciprocal Gating and Attention-fusion Network), a novel deep learning framework specifically designed for smoke removal in robotic surgery workflows. Our approach addresses the unique challenges of surgical smoke-including dense, non-homogeneous distribution and complex light scattering-through a hierarchical encoder-decoder architecture featuring two key innovations: (1) a Dual-Stream Hybrid Attention (DHA) module that combines shifted window attention with frequency-domain processing to capture both local surgical details and global illumination changes, and (2) an Axis-Decomposed Attention (ADA) module that efficiently processes multi-scale features through factorized attention mechanisms. These components are connected via reciprocal cross-gating blocks that enable bidirectional feature modulation between encoder and decoder pathways. Extensive experiments on the DesmokeData and LSD3K surgical datasets demonstrate that RGA-Net achieves superior performance in restoring visual clarity suitable for robotic surgery integration. Our method enhances the surgeon-robot interface by providing consistently clear visualization, laying a technical foundation for alleviating surgeons' cognitive burden, optimizing operation workflows, and reducing iatrogenic injury risks in minimally invasive procedures. These practical benefits could be further validated through future clinical trials involving surgeon usability assessments. The proposed framework represents a significant step toward more reliable and safer robotic surgical systems through computational vision enhancement. △ Less

Submitted 14 February, 2026; originally announced February 2026.

Comments: Accepted by ICRA2026

H.265/HEVC Video Steganalysis Based on CU Block Structure Gradients and IPM Mapping

Authors: Xiang Zhang, Haiyang Xia, Ziwen He, Wenbin Huang, Fei Peng, Zhangjie Fu

Abstract: Existing H.265/HEVC video steganalysis research mainly focuses on detecting the steganography based on motion vectors, intra prediction modes, and transform coefficients. However, there is currently no effective steganalysis method capable of detecting steganography based on Coding Unit (CU) block structure. To address this issue, we propose, for the first time, a H.265/HEVC video steganalysis alg… ▽ More Existing H.265/HEVC video steganalysis research mainly focuses on detecting the steganography based on motion vectors, intra prediction modes, and transform coefficients. However, there is currently no effective steganalysis method capable of detecting steganography based on Coding Unit (CU) block structure. To address this issue, we propose, for the first time, a H.265/HEVC video steganalysis algorithm based on CU block structure gradients and intra prediction mode mapping. The proposed method first constructs a new gradient map to explicitly describe changes in CU block structure, and combines it with a block level mapping representation of IPM. It can jointly model the structural perturbations introduced by steganography based on CU block structure. Then, we design a novel steganalysis network called GradIPMFormer, whose core innovation is an integrated architecture that combines convolutional local embedding with Transformer-based token modeling to jointly capture local CU boundary perturbations and long-range cross-CU structural dependencies, thereby effectively enhancing the capability to perceive CU block structure embedding. Experimental results show that under different quantization parameters and resolution settings, the proposed method consistently achieves superior detection performance across multiple steganography methods based on CU block structure. This study provides a new CU block structure steganalysis paradigm for H.265/HEVC and has significant research value for covert communication security detection. △ Less

Submitted 15 March, 2026; v1 submitted 11 February, 2026; originally announced February 2026.

Embedding Inversion via Conditional Masked Diffusion Language Models

Authors: Han Xiao

Abstract: We frame embedding inversion as conditional masked diffusion, recovering all tokens in parallel through iterative denoising rather than sequential autoregressive generation. A masked diffusion language model is conditioned on the target embedding via adaptive layer normalization, requiring only 8 forward passes with no access to the target encoder at inference time. On 32-token sequences across th… ▽ More We frame embedding inversion as conditional masked diffusion, recovering all tokens in parallel through iterative denoising rather than sequential autoregressive generation. A masked diffusion language model is conditioned on the target embedding via adaptive layer normalization, requiring only 8 forward passes with no access to the target encoder at inference time. On 32-token sequences across three embedding models, the method achieves token recovery through parallel denoising without requiring encoder access, iterative correction, or architecture-specific alignment. Source code and live demo are available at https://github.com/jina-ai/embedding-inversion-demo. △ Less

Submitted 18 February, 2026; v1 submitted 11 February, 2026; originally announced February 2026.

Comments: 8 pages, 3 figures, 4 tables. Code and demo: https://github.com/jina-ai/embedding-inversion-demo

Abstraction Generation for Generalized Planning with Pretrained Large Language Models

Authors: Zhenhe Cui, Huaxiang Xia, Hangjun Shen, Kailun Luo, Yong He, Wei Liang

Abstract: Qualitative Numerical Planning (QNP) serves as an important abstraction model for generalized planning (GP), which aims to compute general plans that solve multiple instances at once. Recent works show that large language models (LLMs) can function as generalized planners. This work investigates whether LLMs can serve as QNP abstraction generators for GP problems and how to fix abstractions via au… ▽ More Qualitative Numerical Planning (QNP) serves as an important abstraction model for generalized planning (GP), which aims to compute general plans that solve multiple instances at once. Recent works show that large language models (LLMs) can function as generalized planners. This work investigates whether LLMs can serve as QNP abstraction generators for GP problems and how to fix abstractions via automated debugging. We propose a prompt protocol: input a GP domain and training tasks to LLMs, prompting them to generate abstract features and further abstract the initial state, action set, and goal into QNP problems. An automated debugging method is designed to detect abstraction errors, guiding LLMs to fix abstractions. Experiments demonstrate that under properly guided by automated debugging, some LLMs can generate useful QNP abstractions. △ Less

Submitted 10 February, 2026; originally announced February 2026.

SAGE: Scalable Agentic 3D Scene Generation for Embodied AI

Authors: Hongchi Xia, Xuan Li, Zhaoshuo Li, Qianli Ma, Jiashu Xu, Ming-Yu Liu, Yin Cui, Tsung-Yi Lin, Wei-Chiu Ma, Shenlong Wang, Shuran Song, Fangyin Wei

Abstract: Real-world data collection for embodied agents remains costly and unsafe, calling for scalable, realistic, and simulator-ready 3D environments. However, existing scene-generation systems often rely on rule-based or task-specific pipelines, yielding artifacts and physically invalid scenes. We present SAGE, an agentic framework that, given a user-specified embodied task (e.g., "pick up a bowl and pl… ▽ More Real-world data collection for embodied agents remains costly and unsafe, calling for scalable, realistic, and simulator-ready 3D environments. However, existing scene-generation systems often rely on rule-based or task-specific pipelines, yielding artifacts and physically invalid scenes. We present SAGE, an agentic framework that, given a user-specified embodied task (e.g., "pick up a bowl and place it on the table"), understands the intent and automatically generates simulation-ready environments at scale. The agent couples multiple generators for layout and object composition with critics that evaluate semantic plausibility, visual realism, and physical stability. Through iterative reasoning and adaptive tool selection, it self-refines the scenes until meeting user intent and physical validity. The resulting environments are realistic, diverse, and directly deployable in modern simulators for policy training. Policies trained purely on this data exhibit clear scaling trends and generalize to unseen objects and layouts, demonstrating the promise of simulation-driven scaling for embodied AI. Code, demos, and the SAGE-10k dataset can be found on the project page here: https://research.nvidia.com/labs/dir/sage/. △ Less

Submitted 20 February, 2026; v1 submitted 10 February, 2026; originally announced February 2026.

Comments: Project Page: https://research.nvidia.com/labs/dir/sage/

SARM: LLM-Augmented Semantic Anchor for End-to-End Live-Streaming Ranking

Authors: Ruochen Yang, Yueyang Liu, Zijie Zhuang, Changxin Lao, Yuhui Zhang, Jiangxia Cao, Jia Xu, Xiang Chen, Haoke Xiao, Xiangyu Wu, Xiaoyou Zhou, Xiao Lv, Shuang Yang, Tingwen Liu, Zhaojie Liu, Han Li, Kun Gai

Abstract: Large-scale live-streaming recommendation requires precise modeling of non-stationary content semantics under strict real-time serving constraints. In industrial deployment, two common approaches exhibit fundamental limitations: discrete semantic abstractions sacrifice descriptive precision through clustering, while dense multimodal embeddings are extracted independently and remain weakly aligned… ▽ More Large-scale live-streaming recommendation requires precise modeling of non-stationary content semantics under strict real-time serving constraints. In industrial deployment, two common approaches exhibit fundamental limitations: discrete semantic abstractions sacrifice descriptive precision through clustering, while dense multimodal embeddings are extracted independently and remain weakly aligned with ranking optimization, limiting fine-grained content-aware ranking. To address these limitations, we propose \textbf{SARM}, an end-to-end ranking architecture that integrates natural-language semantic anchors directly into ranking optimization, enabling fine-grained author representations conditioned on multimodal content. Each semantic anchor is represented as learnable text tokens jointly optimized with ranking features, allowing the model to adapt content descriptions to ranking objectives. A lightweight dual-token gated design captures domain-specific live-streaming semantics, while an asymmetric deployment strategy preserves low-latency online training and serving. Extensive offline evaluation and large-scale A/B tests show consistent improvements over production baselines. SARM is fully deployed and serves over 400 million users daily. △ Less

Submitted 9 February, 2026; originally announced February 2026.

Latent Poincaré Shaping for Agentic Reinforcement Learning

Authors: Hanchen Xia, Baoyou Chen, Zelin Zang, Yutang Ge, Guojiang Zhao, Siyu Zhu

Abstract: We propose LaPha, a method for training AlphaZero-like LLM agents in a Poincaré latent space. Under LaPha, the search process can be visualized as a tree rooted at the prompt and growing outward from the origin toward the boundary of the Poincaré ball, where negative curvature provides exponentially increasing capacity with radius. Using hyperbolic geodesic distance to rule-verified correctness, w… ▽ More We propose LaPha, a method for training AlphaZero-like LLM agents in a Poincaré latent space. Under LaPha, the search process can be visualized as a tree rooted at the prompt and growing outward from the origin toward the boundary of the Poincaré ball, where negative curvature provides exponentially increasing capacity with radius. Using hyperbolic geodesic distance to rule-verified correctness, we define a node potential and assign dense process rewards by potential differences. We further attach a lightweight value head on the same shared latent space, enabling self-guided test-time scaling with almost no additional overhead. On MATH-500, LaPha improves Qwen2.5-Math-1.5B from 66.0% to 88.2%. With value-head-guided search, LaPha-1.5B reaches 56.7% accuracy on AIME'24, and LaPha-7B further achieves 60.0% on AIME'24 and 53.3% on AIME'25. △ Less

Submitted 10 March, 2026; v1 submitted 9 February, 2026; originally announced February 2026.

STEP: Warm-Started Visuomotor Policies with Spatiotemporal Consistency Prediction

Authors: Jinhao Li, Yuxuan Cong, Yingqiao Wang, Hao Xia, Shan Huang, Yijia Zhang, Ningyi Xu, Guohao Dai

Abstract: Diffusion policies have recently emerged as a powerful paradigm for visuomotor control in robotic manipulation due to their ability to model the distribution of action sequences and capture multimodality. However, iterative denoising leads to substantial inference latency, limiting control frequency in real-time closed-loop systems. Existing acceleration methods either reduce sampling steps, bypas… ▽ More Diffusion policies have recently emerged as a powerful paradigm for visuomotor control in robotic manipulation due to their ability to model the distribution of action sequences and capture multimodality. However, iterative denoising leads to substantial inference latency, limiting control frequency in real-time closed-loop systems. Existing acceleration methods either reduce sampling steps, bypass diffusion through direct prediction, or reuse past actions, but often struggle to jointly preserve action quality and achieve consistently low latency. In this work, we propose STEP, a lightweight spatiotemporal consistency prediction mechanism to construct high-quality warm-start actions that are both distributionally close to the target action and temporally consistent, without compromising the generative capability of the original diffusion policy. Then, we propose a velocity-aware perturbation injection mechanism that adaptively modulates actuation excitation based on temporal action variation to prevent execution stall especially for real-world tasks. We further provide a theoretical analysis showing that the proposed prediction induces a locally contractive mapping, ensuring convergence of action errors during diffusion refinement. We conduct extensive evaluations on nine simulated benchmarks and two real-world tasks. Notably, STEP with 2 steps can achieve an average 21.6% and 27.5% higher success rate than BRIDGER and DDIM on the RoboMimic benchmark and real-world tasks, respectively. These results demonstrate that STEP consistently advances the Pareto frontier of inference latency and success rate over existing methods. △ Less

Submitted 8 February, 2026; originally announced February 2026.

Comments: 13 pages, 9 figures

UAV-Enabled Short-Packet Communication via Fluid Antenna Systems

Authors: Xusheng Zhu, Kai-Kit Wong, Hanjiang Hong, Han Xiao, Hao Xu, Tuo Wu, Chan-Byoung Chae

Abstract: This paper develops a framework for analyzing UAV-enabled short-packet communication, leveraging fluid antenna system (FAS)-assisted relaying networks. Operating in the short-packet regime and focusing on challenging urban environments, we derive novel, closed-form expressions for the block error rate (BLER). This is achieved by modeling the spatially correlated Nakagami-$m$ fading link via a trac… ▽ More This paper develops a framework for analyzing UAV-enabled short-packet communication, leveraging fluid antenna system (FAS)-assisted relaying networks. Operating in the short-packet regime and focusing on challenging urban environments, we derive novel, closed-form expressions for the block error rate (BLER). This is achieved by modeling the spatially correlated Nakagami-$m$ fading link via a tractable eigenvalue-based approach. A high-signal-to-noise ratio (SNR) asymptotic analysis is also presented, revealing the system's fundamental diversity order. Building on this analysis, we formulate a novel energy efficiency (EE) maximization problem that, unlike idealized models, uniquely incorporates the non-trivial time and energy overheads of FAS port selection. An efficient hierarchical algorithm is proposed to jointly optimize key system parameters. Numerical results validate our analysis, demonstrating that while FAS provides substantial power gains, the operational overhead creates a critical trade-off. This trade-off dictates an optimal number of FAS ports and a non-trivial optimal UAV deployment altitude, governed by the balance between blockage and path loss. This work provides key insights for FAS-aided UAV communications. △ Less

Submitted 5 February, 2026; originally announced February 2026.

MemGUI-Bench: Benchmarking Memory of Mobile GUI Agents in Dynamic Environments

Authors: Guangyi Liu, Pengxiang Zhao, Yaozhen Liang, Qinyi Luo, Shunye Tang, Yuxiang Chai, Weifeng Lin, Han Xiao, WenHao Wang, Siheng Chen, Zhengxi Lu, Gao Wu, Hao Wang, Liang Liu, Yong Liu

Abstract: Current mobile GUI agent benchmarks systematically fail to assess memory capabilities, with only 5.2-11.8% memory-related tasks and no cross-session learning evaluation. We introduce MemGUI-Bench, a comprehensive memory-centric benchmark with pass@k and staged LLM-as-judge evaluation. Our contributions include: (1) a systematic memory taxonomy analyzing 11 agents across 5 architectures; (2) 128 ta… ▽ More Current mobile GUI agent benchmarks systematically fail to assess memory capabilities, with only 5.2-11.8% memory-related tasks and no cross-session learning evaluation. We introduce MemGUI-Bench, a comprehensive memory-centric benchmark with pass@k and staged LLM-as-judge evaluation. Our contributions include: (1) a systematic memory taxonomy analyzing 11 agents across 5 architectures; (2) 128 tasks across 26 applications where 89.8% challenge memory through cross-temporal and cross-spatial retention; (3) MemGUI-Eval, an automated pipeline with Progressive Scrutiny and 7 hierarchical metrics; and (4) RQ-driven assessment of 11 state-of-the-art agents. Our experiments reveal significant memory deficits across all evaluated systems, identify 5 distinct failure modes, and synthesize 5 actionable design implications. All resources including code, benchmark, and evaluation results will be \textbf{\textit{fully open-sourced and continuously maintained}} at https://lgy0404.github.io/MemGUI-Bench/. △ Less

Submitted 3 February, 2026; originally announced February 2026.

Comments: https://lgy0404.github.io/MemGUI-Bench/

UI-Mem: Self-Evolving Experience Memory for Online Reinforcement Learning in Mobile GUI Agents

Authors: Han Xiao, Guozhi Wang, Hao Wang, Shilong Liu, Yuxiang Chai, Yue Pan, Yufeng Zhou, Xiaoxin Chen, Yafei Wen, Hongsheng Li

Abstract: Online Reinforcement Learning (RL) offers a promising paradigm for enhancing GUI agents through direct environment interaction. However, its effectiveness is severely hindered by inefficient credit assignment in long-horizon tasks and repetitive errors across tasks due to the lack of experience transfer. To address these challenges, we propose UI-Mem, a novel framework that enhances GUI online RL… ▽ More Online Reinforcement Learning (RL) offers a promising paradigm for enhancing GUI agents through direct environment interaction. However, its effectiveness is severely hindered by inefficient credit assignment in long-horizon tasks and repetitive errors across tasks due to the lack of experience transfer. To address these challenges, we propose UI-Mem, a novel framework that enhances GUI online RL with a Hierarchical Experience Memory. Unlike traditional replay buffers, our memory accumulates structured knowledge, including high-level workflows, subtask skills, and failure patterns. These experiences are stored as parameterized templates that enable cross-task and cross-application transfer. To effectively integrate memory guidance into online RL, we introduce Stratified Group Sampling, which injects varying levels of guidance across trajectories within each rollout group to maintain outcome diversity, driving the unguided policy toward internalizing guided behaviors. Furthermore, a Self-Evolving Loop continuously abstracts novel strategies and errors to keep the memory aligned with the agent's evolving policy. Experiments on online GUI benchmarks demonstrate that UI-Mem significantly outperforms traditional RL baselines and static reuse strategies, with strong generalization to unseen applications. Project page: https://ui-mem.github.io △ Less

Submitted 5 February, 2026; originally announced February 2026.

Comments: 23 pages, 16 figures. Project page: https://ui-mem.github.io

BPDQ: Bit-Plane Decomposition Quantization on a Variable Grid for Large Language Models

Authors: Junyu Chen, Jungang Li, Jing Xiong, Wenjie Wang, Qingyao Yang, He Xiao, Zhen Li, Taiqiang Wu, Mengzhao Chen, Zhen Peng, Chaofan Tao, Long Shi, Hongxia Yang, Ngai Wong

Abstract: Large language model (LLM) inference is often bounded by memory footprint and memory bandwidth in resource-constrained deployments, making quantization a fundamental technique for efficient serving. While post-training quantization (PTQ) maintains high fidelity at 4-bit, it deteriorates at 2-3 bits. Fundamentally, existing methods enforce a shape-invariant quantization grid (e.g., the fixed unifor… ▽ More Large language model (LLM) inference is often bounded by memory footprint and memory bandwidth in resource-constrained deployments, making quantization a fundamental technique for efficient serving. While post-training quantization (PTQ) maintains high fidelity at 4-bit, it deteriorates at 2-3 bits. Fundamentally, existing methods enforce a shape-invariant quantization grid (e.g., the fixed uniform intervals of UINT2) for each group, severely restricting the feasible set for error minimization. To address this, we propose Bit-Plane Decomposition Quantization (BPDQ), which constructs a variable quantization grid via bit-planes and scalar coefficients, and iteratively refines them using approximate second-order information while progressively compensating quantization errors to minimize output discrepancy. In the 2-bit regime, BPDQ enables serving Qwen2.5-72B on a single RTX 3090 with 83.85% GSM8K accuracy (vs. 90.83% at 16-bit). Moreover, we provide theoretical analysis showing that the variable grid expands the feasible set, and that the quantization process consistently aligns with the optimization objective in Hessian-induced geometry. Code: github.com/KingdalfGoodman/BPDQ. △ Less

Submitted 3 February, 2026; originally announced February 2026.

Conflict-Resolving and Sharpness-Aware Minimization for Generalized Knowledge Editing with Multiple Updates

Authors: Duy Nguyen, Hanqi Xiao, Archiki Prasad, Elias Stengel-Eskin, Hyunji Lee, Mohit Bansal

Abstract: Large language models (LLMs) rely on internal knowledge to solve many downstream tasks, making it crucial to keep them up to date. Since full retraining is expensive, prior work has explored efficient alternatives such as model editing and parameter-efficient fine-tuning. However, these approaches often break down in practice due to poor generalization across inputs, limited stability, and knowled… ▽ More Large language models (LLMs) rely on internal knowledge to solve many downstream tasks, making it crucial to keep them up to date. Since full retraining is expensive, prior work has explored efficient alternatives such as model editing and parameter-efficient fine-tuning. However, these approaches often break down in practice due to poor generalization across inputs, limited stability, and knowledge conflict. To address these limitations, we propose the CoRSA (Conflict-Resolving and Sharpness-Aware Minimization) training framework, a parameter-efficient, holistic approach for knowledge editing with multiple updates. CoRSA tackles multiple challenges simultaneously: it improves generalization to different input forms and enhances stability across multiple updates by minimizing loss curvature, and resolves conflicts by maximizing the margin between new and prior knowledge. Across three widely used fact editing benchmarks, CoRSA achieves significant gains in generalization, outperforming baselines with average absolute improvements of 12.42% over LoRA and 10% over model editing methods. With multiple updates, it maintains high update efficacy while reducing catastrophic forgetting by 27.82% compared to LoRA. CoRSA also generalizes to the code domain, outperforming the strongest baseline by 5.48% Pass@5 in update efficacy. △ Less

Submitted 3 February, 2026; originally announced February 2026.

Comments: 22 pages, 8 figures. Code link: https://github.com/duykhuongnguyen/CoRSA

Acceleration of Atomistic NEGF: Algorithms, Parallelization, and Machine Learning

Authors: Mathieu Luisier, Nicolas Vetsch, Alexander Maeder, Vincent Maillou, Anders Winka, Leonard Deuschle, Chen Hao Xia, Manasa Kaniselvan, Marko Mladenovic, Jiang Cao, Alexandros Nikolaos Ziogas

Abstract: The Non-equilibrium Green's function (NEGF) formalism is a particularly powerful method to simulate the quantum transport properties of nanoscale devices such as transistors, photo-diodes, or memory cells, in the ballistic limit of transport or in the presence of various scattering sources such as electronphonon, electron-photon, or even electron-electron interactions. The inclusion of all these m… ▽ More The Non-equilibrium Green's function (NEGF) formalism is a particularly powerful method to simulate the quantum transport properties of nanoscale devices such as transistors, photo-diodes, or memory cells, in the ballistic limit of transport or in the presence of various scattering sources such as electronphonon, electron-photon, or even electron-electron interactions. The inclusion of all these mechanisms has been first demonstrated in small systems, composed of a few atoms, before being scaled up to larger structures made of thousands of atoms. Also, the accuracy of the models has kept improving, from empirical to fully ab-initio ones, e.g., density functional theory (DFT). This paper summarizes key (algorithmic) achievements that have allowed us to bring DFT+NEGF simulations closer to the dimensions and functionality of realistic systems. The possibility of leveraging graph neural networks and machine learning to speed up ab-initio device simulations is discussed as well. △ Less

Submitted 3 February, 2026; originally announced February 2026.

Full end-to-end diagnostic workflow automation of 3D OCT via foundation model-driven AI for retinal diseases

Authors: Jinze Zhang, Jian Zhong, Li Lin, Jiaxiong Li, Ke Ma, Naiyang Li, Meng Li, Yuan Pan, Zeyu Meng, Mengyun Zhou, Shang Huang, Shilong Yu, Zhengyu Duan, Sutong Li, Honghui Xia, Juping Liu, Dan Liang, Yantao Wei, Xiaoying Tang, Jin Yuan, Peng Xiao

Abstract: Optical coherence tomography (OCT) has revolutionized retinal disease diagnosis with its high-resolution and three-dimensional imaging nature, yet its full diagnostic automation in clinical practices remains constrained by multi-stage workflows and conventional single-slice single-task AI models. We present Full-process OCT-based Clinical Utility System (FOCUS), a foundation model-driven framework… ▽ More Optical coherence tomography (OCT) has revolutionized retinal disease diagnosis with its high-resolution and three-dimensional imaging nature, yet its full diagnostic automation in clinical practices remains constrained by multi-stage workflows and conventional single-slice single-task AI models. We present Full-process OCT-based Clinical Utility System (FOCUS), a foundation model-driven framework enabling end-to-end automation of 3D OCT retinal disease diagnosis. FOCUS sequentially performs image quality assessment with EfficientNetV2-S, followed by abnormality detection and multi-disease classification using a fine-tuned Vision Foundation Model. Crucially, FOCUS leverages a unified adaptive aggregation method to intelligently integrate 2D slices-level predictions into comprehensive 3D patient-level diagnosis. Trained and tested on 3,300 patients (40,672 slices), and externally validated on 1,345 patients (18,498 slices) across four different-tier centers and diverse OCT devices, FOCUS achieved high F1 scores for quality assessment (99.01%), abnormally detection (97.46%), and patient-level diagnosis (94.39%). Real-world validation across centers also showed stable performance (F1: 90.22%-95.24%). In human-machine comparisons, FOCUS matched expert performance in abnormality detection (F1: 95.47% vs 90.91%) and multi-disease diagnosis (F1: 93.49% vs 91.35%), while demonstrating better efficiency. FOCUS automates the image-to-diagnosis pipeline, representing a critical advance towards unmanned ophthalmology with a validated blueprint for autonomous screening to enhance population scale retinal care accessibility and efficiency. △ Less

Submitted 3 February, 2026; originally announced February 2026.

Privasis: Synthesizing the Largest "Public" Private Dataset from Scratch

Authors: Hyunwoo Kim, Niloofar Mireshghallah, Michael Duan, Rui Xin, Shuyue Stella Li, Jaehun Jung, David Acuna, Qi Pang, Hanshen Xiao, G. Edward Suh, Sewoong Oh, Yulia Tsvetkov, Pang Wei Koh, Yejin Choi

Abstract: Research involving privacy-sensitive data has always been constrained by data scarcity, standing in sharp contrast to other areas that have benefited from data scaling. This challenge is becoming increasingly urgent as modern AI agents--such as OpenClaw and Gemini Agent--are granted persistent access to highly sensitive personal information. To tackle this longstanding bottleneck and the rising ri… ▽ More Research involving privacy-sensitive data has always been constrained by data scarcity, standing in sharp contrast to other areas that have benefited from data scaling. This challenge is becoming increasingly urgent as modern AI agents--such as OpenClaw and Gemini Agent--are granted persistent access to highly sensitive personal information. To tackle this longstanding bottleneck and the rising risks, we present Privasis (i.e., privacy oasis), the first million-scale fully synthetic dataset entirely built from scratch--an expansive reservoir of texts with rich and diverse private information--designed to broaden and accelerate research in areas where processing sensitive social data is inevitable. Compared to existing datasets, Privasis, comprising 1.4 million records, offers orders-of-magnitude larger scale with quality, and far greater diversity across various document types, including medical history, legal documents, financial records, calendars, and text messages with a total of 55.1 million annotated attributes such as ethnicity, date of birth, workplace, etc. We leverage Privasis to construct a parallel corpus for text sanitization with our pipeline that decomposes texts and applies targeted sanitization. Our compact sanitization models (<=4B) trained on this dataset outperform state-of-the-art large language models, such as GPT-5 and Qwen-3 235B. We plan to release data, models, and code to accelerate future research on privacy-sensitive domains and agents. △ Less

Submitted 3 February, 2026; originally announced February 2026.

Comments: For code and data, see https://privasis.github.io

TF-Lane: Traffic Flow Module for Robust Lane Perception

Authors: Yihan Xie, Han Xia, Zhen Yang

Abstract: Autonomous driving systems require robust lane perception capabilities, yet existing vision-based detection methods suffer significant performance degradation when visual sensors provide insufficient cues, such as in occluded or lane-missing scenarios. While some approaches incorporate high-definition maps as supplementary information, these solutions face challenges of high subscription costs and… ▽ More Autonomous driving systems require robust lane perception capabilities, yet existing vision-based detection methods suffer significant performance degradation when visual sensors provide insufficient cues, such as in occluded or lane-missing scenarios. While some approaches incorporate high-definition maps as supplementary information, these solutions face challenges of high subscription costs and limited real-time performance. To address these limitations, we explore an innovative information source: traffic flow, which offers real-time capabilities without additional costs. This paper proposes a TrafficFlow-aware Lane perception Module (TFM) that effectively extracts real-time traffic flow features and seamlessly integrates them with existing lane perception algorithms. This solution originated from real-world autonomous driving conditions and was subsequently validated on open-source algorithms and datasets. Extensive experiments on four mainstream models and two public datasets (Nuscenes and OpenLaneV2) using standard evaluation metrics show that TFM consistently improves performance, achieving up to +4.1% mAP gain on the Nuscenes dataset. △ Less

Submitted 1 February, 2026; originally announced February 2026.

Comments: 9 pages, 7 figures, 7 tables

ACM Class: I.4.8

Embedding Compression via Spherical Coordinates

Authors: Han Xiao

Abstract: We present an $ε$-bounded compression method for unit-norm embeddings that achieves 1.5$\times$ compression, 25% better than the best prior lossless method. The method exploits that spherical coordinates of high-dimensional unit vectors concentrate around $π/2$, causing IEEE 754 exponents to collapse to a single value and high-order mantissa bits to become predictable, enabling entropy coding of b… ▽ More We present an $ε$-bounded compression method for unit-norm embeddings that achieves 1.5$\times$ compression, 25% better than the best prior lossless method. The method exploits that spherical coordinates of high-dimensional unit vectors concentrate around $π/2$, causing IEEE 754 exponents to collapse to a single value and high-order mantissa bits to become predictable, enabling entropy coding of both. Reconstruction error is bounded by float32 machine epsilon ($1.19 \times 10^{-7}$), making reconstructed values indistinguishable from originals at float32 precision. Evaluation across 26 configurations spanning text, image, and multi-vector embeddings confirms consistent compression improvement with zero measurable retrieval degradation on BEIR benchmarks. △ Less

Submitted 25 March, 2026; v1 submitted 21 January, 2026; originally announced February 2026.

Comments: Accepted at ICLR 2026 Workshop on Geometry-grounded Representation Learning and Generative Modeling (GRaM). 13 pages, 2 figures. Code: https://github.com/jina-ai/jzip

MSC Class: 68T50 ACM Class: I.2.7

AutoBinder Agent: An MCP-Based Agent for End-to-End Protein Binder Design

Authors: Fukang Ge, Jiarui Zhu, Linjie Zhang, Haowen Xiao, Xiangcheng Bao, Fangnan Xie, Danyang Chen, Yanrui Lu, Yuting Wang, Ziqian Guan, Lin Gu, Jinhao Bi, Yingying Zhu

Abstract: Modern AI technologies for drug discovery are distributed across heterogeneous platforms-including web applications, desktop environments, and code libraries-leading to fragmented workflows, inconsistent interfaces, and high integration overhead. We present an agentic end-to-end drug design framework that leverages a Large Language Model (LLM) in conjunction with the Model Context Protocol (MCP) t… ▽ More Modern AI technologies for drug discovery are distributed across heterogeneous platforms-including web applications, desktop environments, and code libraries-leading to fragmented workflows, inconsistent interfaces, and high integration overhead. We present an agentic end-to-end drug design framework that leverages a Large Language Model (LLM) in conjunction with the Model Context Protocol (MCP) to dynamically coordinate access to biochemical databases, modular toolchains, and task-specific AI models. The system integrates four state-of-the-art components: MaSIF (MaSIF-site and MaSIF-seed-search) for geometric deep learning-based identification of protein-protein interaction (PPI) sites, Rosetta for grafting protein fragments onto protein backbones to form mini proteins, ProteinMPNN for amino acid sequences redesign, and AlphaFold3 for near-experimental accuracy in complex structure prediction. Starting from a target structure, the framework supports de novo binder generation via surface analysis, scaffold grafting and pose construction, sequence optimization, and structure prediction. Additionally, by replacing rigid, script-based workflows with a protocol-driven, LLM-coordinated architecture, the framework improves reproducibility, reduces manual overhead, and ensures extensibility, portability, and auditability across the entire drug design process. △ Less

Submitted 16 January, 2026; originally announced February 2026.

Comments: 4 pages, 3 figures

DELNet: Continuous All-in-One Weather Removal via Dynamic Expert Library

Authors: Shihong Liu, Kun Zuo, Hanguang Xiao

Abstract: All-in-one weather image restoration methods are valuable in practice but depend on pre-collected data and require retraining for unseen degradations, leading to high cost. We propose DELNet, a continual learning framework for weather image restoration. DELNet integrates a judging valve that measures task similarity to distinguish new from known tasks, and a dynamic expert library that stores expe… ▽ More All-in-one weather image restoration methods are valuable in practice but depend on pre-collected data and require retraining for unseen degradations, leading to high cost. We propose DELNet, a continual learning framework for weather image restoration. DELNet integrates a judging valve that measures task similarity to distinguish new from known tasks, and a dynamic expert library that stores experts trained on different degradations. For new tasks, the valve selects top-k experts for knowledge transfer while adding new experts to capture task-specific features; for known tasks, the corresponding experts are directly reused. This design enables continuous optimization without retraining existing models. Experiments on OTS, Rain100H, and Snow100K demonstrate that DELNet surpasses state-of-the-art continual learning methods, achieving PSNR gains of 16\%, 11\%, and 12\%, respectively. These results highlight the effectiveness, robustness, and efficiency of DELNet, which reduces retraining cost and enables practical deployment in real-world scenarios. △ Less

Submitted 30 January, 2026; originally announced January 2026.

Comments: Accepted by the ICASSP conference, not yet officially published