arXiv:2507.01932v2 Announce Type: replace-cross
Abstract: We study a class of nonconvex-nonconcave minimax problems in which the inner maximization problem satisfies a local Kurdyka-Lojasiewicz (KL) condition that may vary with the outer minimization variable. In contrast to the global KL or Polyak-Lojasiewicz (PL) conditions commonly assumed in the literature -- which are significantly stronger and often too restrictive in practice -- this local KL condition accommodates a broader range of practical scenarios. However, it also introduces new analytical challenges. In particular, as an optimization algorithm progresses toward a stationary point of the problem, the region over which the KL condition holds may shrink, resulting in a more intricate and potentially ill-conditioned landscape. To address this challenge, we show that the associated maximal function is locally generalized H\"older smooth. Leveraging this key property, we develop an inexact proximal gradient method for solving the minimax problem, where the inexact gradient of the maximal function is computed by applying a proximal gradient method to a KL-structured subproblem. Under mild assumptions, we establish complexity guarantees for computing an approximate stationary point of the minimax problem.
Science Journals
arXiv:2605.13318v2 Announce Type: replace
Abstract: Chatbot usage has increased, including in fields for which they were never developed for--notably mental health support. To that end, we introduce Validations of Ethical and Responsible AI in Mental Health (VERA-MH), a novel clinically-validated evaluation for safety of chatbots in the context of mental health support. The first iteration of VERA-MH focuses on Suicidal Ideation (SI) risks, by assessing how well chatbots can responds to users that might be in crisis.
VERA-MH is comprised of three steps: conversation simulation, conversation judging and model rating. First, to simulate conversations with the chatbot under evaluation, another chatbot is tasked with role-playing users based on specific personas. Such user personas have been developed under clinical guidance, to make sure that, among others, multiple risk factors, demographic characteristics and disclosure factors were represented. In the judging step, a second support model is used as an LLM-as-a-Judge, together with a clinically-developed rubric. The rubric is structured as a flow, with a single Yes/No question asked each time, to improve answers' consistency and highlight models' failure modes. In the last stage, results of each conversation are aggregated to present the final evaluation of the chatbot. Together with the framework, we present the result of the evaluations for four leading LLM providers.
arXiv:2605.19492v1 Announce Type: new
Abstract: In the context of building acoustics, sound transmission loss estimations are crucial to quantify the noise pollution in buildings. When developing building prototypes in the sense of an acoustic-oriented design process, it is desirable to have an virtual prototype, especially in early development stages, to estimate, for instance, the influence of different material or geometry configurations on to the sound transmission loss. This contribution aims to present a simple virtual prototype of an building acoustics test facility in accordance with DIN EN ISO 10140 for the measurement of the sound transmission loss of single- and double-leaf walls with and without insulation. Here, the finite element method is used as the numerical modelling method of choice. In the course of this, geometry and mesh creation was done using SALOME 9.14 whereas the institute's in-house research code elPaSo was utilised for the matrix assembly and solving procedure. At first, elPaSo was verified by the commercial software COMSOL 6.3 considering a small-scale test facility. Afterwards, the large-scale test facility finite element model was created using a frequency- and domain-specific discretisation approach. The sound transmission loss of three different test specimens was estimated in one-third-octave bands from 8 Hz to 630 Hz, where the double-leaf wall with insulation exhibited good agreement to the theoretical sound transmission loss profile from literature.
arXiv:2605.19096v1 Announce Type: new
Abstract: Two widely used randomized algorithms are the sketch-and-solve method for least-squares regression and the randomized SVD for low-rank approximation. These algorithms apply a random embedding to compress a target matrix, and they perform computations on the compressed matrix to save computational cost. This paper asks, what is the optimal random embedding in these algorithms? Also, what is the sharpest possible error bound for the optimal embedding? The paper proves that a random orthonormal matrix is minimax optimal for the sketch-and-solve algorithm while any rotation-invariant embedding is minimax optimal for the randomized SVD. Following these results, the paper obtains the best possible error bounds for sketched least-squares and the randomized SVD. Last, empirical experiments provide evidence of universality phenomena, in which several random embeddings lead to similar accuracy to the optimal embeddings in practice.
arXiv:2605.15846v2 Announce Type: replace
Abstract: Coding agents are increasingly deployed in real software development, where a single version iteration requires months of coordinated work across many files. However, most existing benchmarks focus predominantly on single-issue bug fixes from Python repositories, with coarse pass/fail evaluation outcomes, and thus fail to capture long-horizon, multi-target development at real engineering scale. To address this gap, we present RoadmapBench, a benchmark of 115 long-horizon coding tasks grounded in real open-source version upgrades across 17 repositories and 5 programming languages. Each task places the agent on a source-version code snapshot and provides a multi-target roadmap instruction requiring it to implement the functionality introduced in the target version, with a median modification of 3,700 lines across 51 files. We conduct a systematic evaluation on thirteen frontier models and find that even the strongest, Claude-Opus-4.7, resolves only 39.1% of tasks, while the weakest achieves merely 5.2%, in stark contrast to existing bug-fix benchmarks, suggesting that long-horizon software development remains a largely unsolved problem.
arXiv:2310.11203v2 Announce Type: replace
Abstract: We introduce a novel strategy to train randomised predictors in federated learning, where each node of the network aims at preserving its privacy by releasing a local predictor but keeping secret its training dataset with respect to the other nodes.
We then build a global randomised predictor which inherits the properties of the local private predictors in the sense of a PAC-Bayesian generalisation bound. We consider the synchronous case where all nodes share the same training objective (derived from a generalisation bound), and the heterogenous and homogenous cases where each node may have its own personalised training objective. We show through a series of numerical experiments that our approach achieves a comparable predictive performance to that of the batch approach where all datasets are shared across nodes. Moreover the predictors are supported by numerically nonvacuous generalisation bounds while preserving privacy for each node. We explicitly compute the increment on predictive performance and generalisation bounds for our two federated settings, highlighting the price to pay to preserve privacy.
arXiv:2605.19554v1 Announce Type: new
Abstract: Instilling creativity in text-to-image (T2I) generation presents a significant challenge, as it requires synthesized images to exhibit not only visual novelty and surprise, but also artistic value. Current T2I models, however, are largely optimized for literal text-image alignment with their data distribution, and their noise prediction networks constrain the generation to high-probability regions, consequently generating outputs that lack authentic creativity. To address this, we propose a Self-Creative Diffusion (SCDiff) model for meaningful T2I generations featuring two core modules: a learnable spatial weighting (LSW) module and a visual-semantic mixing loss (VSML). The LSW module designs a parametric Kaiser-Bessel window to reinforce central image features, fostering novel and surprising generation. The VSML module introduces a dual loss function: a similarity loss constrains that the new images align with its textual description, while a diversity loss maximizes its distinction from the original image, enhancing both semantic value and visual novelty. Extensive experiments demonstrate that our model substantially improves creativity, semantic alignment, and visual coherence, offering a simple yet powerful framework for generating creative objects.
arXiv:2605.19974v1 Announce Type: new
Abstract: The generation of immersive and navigable 3D environments is increasingly prevalent with the growing adoption of virtual reality and 3D content. However, recent methods face a fundamental limitation: they cannot produce 3D worlds that simultaneously (i) are navigable over long-range spatial extents and (ii) cover the complete omnidirectional field of view ($360^\circ$ horizontally and $180^\circ$ vertically). To address this challenge, we introduce SphericalDreamer, a method for generating fully immersive and long-range 3D outdoor environments from textual prompts. Our approach is built on the generation of multiple panoramic images, which are subsequently lifted into 3D and fused together while maintaining visual and geometric consistency. SphericalDreamer produces highly detailed, fully immersive 3D environments, while substantially improving scale and navigability compared to prior approaches.
arXiv:2605.16820v2 Announce Type: replace
Abstract: Reasoning about array data structures is a key requirement for many applications in hardware and software verification, especially in combination with machine integers. The Satisfiability Modulo Theories (SMT) theory of extensional arrays provides array read and write operators and allows extensionality over arrays. This is sufficient to express many aspects of computer-aided verification, but lacks succinctness to efficiently deal with arrays that are initialized with a default value. Existing procedures for extending the SMT-LIB theory of arrays with support for constant arrays are limited to arrays with infinite index domains, and existing implementations in SMT solvers only support a fragment of the theory for finite index domains. In this paper, we present a novel decision procedure for the theory of arrays with constant arrays that supports arbitrary index domains and is not limited to the infinite case. We present our procedure as an abstract calculus and show its refutational and satisfiability soundness. We implement a decision procedure based on our calculus in the state-of-the-art SMT solver Bitwuzla and evaluate its performance on a diverse collection of benchmarks and use cases.
arXiv:2511.16316v2 Announce Type: replace
Abstract: Integrated sensing and communication (ISAC) will be central to 6G-enabled transportation, providing both seamless connectivity and high-precision sensing. However, this tight integration exposes attack points not encountered in pure sensing and communication systems. In this article, we identify unique ISAC-induced security challenges and opportunities in three interrelated domains: cyber-physical (where manipulation of sensors and actuators can mislead perception and control), physical-layer (where over-the-air signals are vulnerable to spoofing and jamming) and protocol (where complex cryptographic protocols cannot detect lower-layer attacks). Building on these insights, we put forward a multi-domain security vision for 6G transportation and propose an integrated security framework that unifies protection across domains by leveraging existing ISAC measurements for lightweight cross-checks.
arXiv:2605.19562v1 Announce Type: new
Abstract: This paper presents a learning-augmented trajectory planning framework for cooperative unmanned aerial vehicle (UAV) and unmanned ground vehicle (UGV) handover missions. While centralized trajectory optimization ensures dynamic feasibility and task optimality, its high computational cost limits real-time applicability. We propose a neural surrogate planner utilizing decoupled encoder-decoder long short-term memory (LSTM) networks to generate coordinated handover trajectory predictions from the task specifications. These predictions serve as informed warm starts for the downstream centralized optimizer, thereby accelerating convergence to dynamically feasible solutions. Benchmark evaluations demonstrate that the learning-augmented planning framework achieves more than a threefold speedup and 100% optimization success rate compared to cold start optimization. The results indicate that combining data-driven inference with model-based refinement enables fast and reliable trajectory generation for heterogeneous multi-robot systems.
arXiv:2605.16137v2 Announce Type: replace
Abstract: Generating simulation-ready tabletop scenes from task instructions is an intriguing and promising research direction in the field of Embodied AI. However, existing task-to-scene generation methods rely exclusively on large language models (LLMs) to predict scene layouts, inevitably yielding object collisions or floating due to LLMs' inherent limitations in 3D spatial reasoning. In this paper, we present STABLE, a semantics-physics dual-system tailored for simulation-ready tabletop scene generation. STABLE consists of two complementary modules: (i) a Semantic Reasoner, a fine-tuned LLM trained on a structured tabletop scene dataset to generate coarse layouts from input task instructions, and (ii) a Physics Corrector, a physics-aware flow-based denoising model that outputs pose updates to refine layouts, which ensures the physical plausibility of scenes while preserves semantic alignment with task instructions. STABLE adopts a progressive generation paradigm: by alternating between the Semantic Reasoner and Physics Corrector, it incrementally expands the scene from task-critical objects to background objects. Experiments demonstrate that STABLE successfully generates simulation-ready tabletop scenes that strictly conform to task instructions and significantly enhances the physical validity of scenes over prior art.
arXiv:2602.02513v2 Announce Type: replace
Abstract: Artificial intelligence has shown remarkable success in materials discovery and property prediction, particularly for crystalline and polymer systems where material properties and structures are dominated by discrete graph representations. Such graph-central paradigm breaks down on composite materials, which possess continuous and nonlinear design spaces. General composite descriptors, e.g., fiber volume and misalignment angle, cannot fully capture the fiber distributions that determine microstructural characteristics, necessitating the integration of heterogeneous data sources through multimodal learning. Existing alignment-oriented frameworks have proven effective on abundant crystal or polymer data under discrete, unique graph-property mapping assumptions, but fail to address the highly continuous composite design space under extreme data scarcity. In this work we introduce ORDinal-aware imagE-tabulaR alignment (ORDER), a multimodal pretraining framework that establishes ordinality as a core principle for material representations. ORDER ensures that materials with similar target properties occupy nearby regions in the latent space, which effectively preserves the continuous nature of composite properties and enables meaningful interpolation between sparsely observed designs. We evaluate ORDER on a Nanofiber-reinforced composite dataset and a carbon fiber T700 dataset. ORDER and its variants outperform both alignment-oriented and customized property-aware contrastive baselines across property prediction, cross-modal retrieval, and microstructure generation tasks. We further introduce physics-based ordinal surrogate signals avoiding the need for full property annotation during pretrain. Our work demonstrates learning continuous multimodal features are fundamental for composite materials, and provides a reliable pathway toward data-efficient universal multimodal intelligent systems.
arXiv:2605.19802v1 Announce Type: cross
Abstract: Data-centric materials science is changing how materials are discovered, optimized, manufactured, and qualified, yet many deployment-limiting materials problems still depend on experimental, processing-rich, device-level, and field-relevant data that are difficult to capture in conventional materials databases. This perspective argues that the Great Plains and adjacent interior research corridor can make a distinctive national contribution by organizing distributed experimental assets into a trusted regional materials-data ecosystem. The proposed model emphasizes FAIR metadata, provenance, persistent sample identifiers, uncertainty-aware modeling, semi-closed-loop workflows, stackable workforce training, and tiered governance for academic, public, controlled-access, and industry-protected data. We identify five coupled barriers -- fragmented data, weak algorithm--laboratory translation, uneven access to cyberinfrastructure and technical staff, workforce gaps at the materials--data interface, and insufficient incentives for sharing and reuse -- and propose a staged roadmap for addressing them. A high-purity germanium pilot illustrates how regional strengths can be converted into reusable datasets, benchmark models, trained personnel, and decision-improving workflows. The broader message is that regional leadership in data-centric materials science will depend less on geographic concentration than on trustworthy data practices, interoperable infrastructure, cross-trained people, and application-driven materials challenges.
arXiv:2605.20118v1 Announce Type: new
Abstract: High-energy (mJ) and high-peak-power (MW) multicycle terahertz (THz) pulses are essential for nonlinear THz spectroscopy and compact accelerator technologies, yet their generation by nonlinear optical frequency conversion remains inefficient and imposes severe demands on femtosecond driving lasers. Amplifying existing THz pulses offers an appealing alternative, but no power-scalable amplifier has been realized in the sub-THz regime. Here, we demonstrate an all-optical THz amplifier operating at 0.35 THz based on the modulation of nanosecond laser pulses by a weak THz field in periodically poled lithium niobate (PPLN). The THz-induced phase modulation is converted into an amplitude modulation using controlled group delay dispersion, forming a tailored pulse train that can efficiently drive high-energy THz generation in a second crystal, thereby amplifying the THz seed. By analogy to electronic klystrons, we term this device the Phlystron, in which the electron beam carrying the power is replaced by a photon beam. In this proof-of-concept experiment, a 3.3-fold increase in THz energy is achieved with commercial crystals. Scaling analysis indicates the potential for higher gain when using large-aperture PPLN devices and multi-stage amplification. The Phlystron thus provides a scalable route to powerful multicycle THz sources driven by readily available narrowband lasers.
arXiv:2605.19568v1 Announce Type: new
Abstract: Embedding models are pivotal in industrial information retrieval systems like search and advertising. However, existing pretrained models often exhibit fixed architectures and embedding dimensionalities, posing significant challenges when adapting them to diverse deployment scenarios with varying business-driven constraints. A common practice involves fine-tuning with partial parameter initialization from larger pretrained models for resource-constrained tasks. This method is often suboptimal as the misalignment between pretraining and downstream usage prevents full realization of pretraining benefits. To address this limitation, we introduce m3BERT: a Modern, Multi-lingual, Matryoshka Bidirectional Encoder, which features a novel pretraining strategy that jointly optimizes representations across both transformer layers and multiple embedding dimensions. This enables a single model to be tailored to varied resource and accuracy targets while maintaining consistency with pretraining. Incorporating recent architectural improvements, m3BERT uses a three-stage pretraining: monolingual pretraining, multilingual adaptation to serve diverse user bases, and crucial continual pretraining on a massive web domain corpus to enhance utility in commercial retrieval. m3BERT significantly outperforms state-of-the-art embedding models in Bing-Click, a large-scale industrial retrieval dataset, showcasing its practical versatility as an efficient foundation for resource-aware industrial retrieval systems. Further experiments on public datasets also confirm the general effectiveness of our multigranular Matryoshka pretraining strategy.
arXiv:2605.19446v1 Announce Type: new
Abstract: Recently, pre-trained encoders have gained widespread use due to their strong capability in representation extraction. However, they are vulnerable to downstream-agnostic attacks (DAAs). Existing DAA methods operate under a permissive threat model, where an attack is successful if the generated downstream-agnostic adversarial examples (DAEs) change the original prediction, without requiring a specific target. In this paper, we propose a Targeted DAA (TDAA) method under a stricter threat model requiring the attack to be both targeted and downstream-agnostic. Since the downstream task is unknown and encoders do not directly produce predictions, achieving a targeted attack is particularly challenging. To address this, we introduce a novel component termed the 'threat image', pre-selected by the attacker as the target. Specifically, a generator is designed to produce example-specific adversarial perturbations that compel the victim encoder to output identical features for both the DAEs and the threat image. Unlike previous DAA methods that generate a single shared perturbation for all samples, which often fails due to image diversity, our method adopts an example-specific paradigm. This generates tailored perturbations for each image to ensure a high attack success rate and invisibility. By leveraging the threat image as a feature-level anchor, our method builds a task-agnostic bridge to reveal the vulnerabilities of the victim encoder. Extensive experiments on 10 self-supervised methods across 3 benchmark datasets demonstrate the effectiveness of our approach and reveal the pronounced vulnerability of pre-trained encoders. The code will be made publicly available after the review period.
arXiv:2605.18777v1 Announce Type: new
Abstract: Mapping large origin-destination (OD) datasets remains challenging because flow maps become cluttered, meaningful patterns occur at multiple spatial scales, and existing flow-mapping approaches frequently rely on predefined aggregation units or manual generalization. This paper presents XFlowMap, a framework for the cross-scale generalization and mapping of massive OD data. Specifically, the framework integrates cross-scale flow pattern (cluster) detection, automated flow map generalization, and a new cartographic representation for analyzing and visualizing complex origin-destination flow structures. The approach detects salient flow patterns at their appropriate origin and destination scales, extracts high-level structures, and generates a new flow map representation that supports holistic interpretation of complex origin-destination flow patterns. A scan-statistic-based procedure is developed to evaluate and generalize cross-scale flow clusters. The detected clusters are then visualized using a novel flow symbol that integrates location, direction, strength, and OD scales in a single representation. The framework supports both area-based and point-based OD data, is robust to sparse and noisy datasets, and enables comparative mapping of stratified flow data. Experiments with synthetic data and U.S. migration data demonstrate that the method effectively extracts meaningful cross-scale flow patterns and produces clear, information-rich flow maps for large mobility datasets, supporting both static presentation and interactive exploration.
arXiv:2605.19936v1 Announce Type: new
Abstract: Has the style of scientific communication changed due to the growing use of large language models in the writing process? We address this question in the domain of Natural Language Processing by leveraging two data resources we create: a naturalistic corpus of over 37,000 papers from the ACL Anthology (2020-2024); and a synthetic dataset of 3,000 human-written passages and their LLM-generated improvements. We first implement a series of diachronic lexical analyses, showing that both word frequency and usage contexts have changed significantly over time, indicating semantic specialization in some cases and generalization in others. Broadening our perspective, we then model a range of more complex stylistic features and find that LLM-modified texts more frequently contain certain syntactic constructions, more complex and longer words and a lower lexical diversity. Finally, we connect these changes in writing practices to subjective reading experience through a pilot annotation study with 20 domain experts. They overall rate LLM-improved texts as more understandable and exciting, but also express negative qualitative attitudes towards LLMs, highlighting the strongly subjective effect of AI-assisted writing on reading experience.
arXiv:2605.19597v1 Announce Type: new
Abstract: Evaluating large language models (LLMs) on natural-language logical reasoning is essential because rule-governed tasks require conclusions to follow strictly from stated premises. Many existing logical-reasoning benchmarks are generated by templating natural-language items from sampled formulas, provide only coarse or unaudited formal annotations, and are now quickly saturated by frontier reasoning models. We present LLMEval-Logic, a Chinese logical reasoning benchmark built from realistic situational scenarios. Its pipeline forward-authors and expert-audits natural-language items together with their reference formalizations, verifies annotated answers with Z3, constructs expert rubrics for natural-to-formal grading, and hardens selected items through a closed-loop adversarial workflow. The benchmark is released in two paired subsets: a 246-item Base subset shipped with 1,400 expert-developed rubric atoms, and a 190-item Hard subset with 938 multi-step sub-questions over closed model spaces. Evaluating 14 frontier LLMs on LLMEval-Logic reveals substantial gaps in current models: the best model reaches only 37.5% Hard Item Accuracy, and even with reference symbols the highest joint Z3+Rubric formalization score among evaluated models reaches only 60.16%. Our benchmark is publicly available at https://github.com/llmeval/LLMEval-Logic.
arXiv:2605.20054v1 Announce Type: new
Abstract: Treating syntactic equality as a logical connective -- governed by left- and right-introduction rules within the sequent calculus -- offers an elegant and powerful approach to term identity. This treatment of equality allows for the derivation of core mathematical principles, such as Peano's axioms (excluding induction), and serves as a foundation for the Abella interactive proof assistant. However, integrating this equality into automated proof search remains challenging. We present a proof search procedure that extends unification to handle the complexities of quantifier alternation and equations that occur in both positive and negative occurrences. While established logical frameworks such as $\lambda$Prolog and LF lack direct support for this kind of equality, our procedure enables a lightweight logical framework that addresses this gap. Our system enables unification-aware proof search across a diverse range of first-order sequent calculi that can directly use this form of equality.
arXiv:2605.20108v1 Announce Type: new
Abstract: While conventional (k=1) discrete-time barrier certificate conditions impose strict safety constraints by requiring the function to be non-increasing at every step, k-inductive barrier certificates relax this by allowing a temporary increase -- up to k-1 times, each within a threshold $\epsilon$ -- while maintaining overall safety, and improving flexibility. This paper leverages neural networks and constructs k-inductive neural barrier certificates (k-NBCs) for (partially) unknown nonlinear systems. While neural networks offer scalability in the design process, they lack formal guarantees, requiring additional approaches such as counterexample-guided inductive synthesis (CEGIS) with satisfiability modulo theories (SMT) for verification. However, the CEGIS-SMT framework requires knowledge of system dynamics, which is unavailable in practical settings. To address this, we leverage the generalization of the Willems et al.'s fundamental lemma, using a single state trajectory, to construct a data-driven representation of (partially) unknown models for SMT verification without sacrificing accuracy. Additionally, CEGIS-SMT further removes the constraint of restricting barrier certificates to specific function classes, such as sum-of-squares, enabling greater flexibility in their design. We validate our approach on three nonlinear case studies with (partially) unknown dynamics.
arXiv:2605.19559v1 Announce Type: new
Abstract: The rapid development of Multimodal Large Language Models (MLLMs) has led to growing interest in egocentric video understanding, specifically the ability for MLLMs to recognize fine-grained hand-object interactions, track object state changes over time, and reason about manipulative processes in dynamic environments from a first-person perspective. However, existing egocentric video benchmarks suffer from \textbf{limited grounded rationale evaluation}, offering limited support for fine-grained operation-centric reasoning and rarely examining whether model rationales are grounded in explicit spatio-temporal evidence. To address this gap, we introduce \textbf{EgoCoT-Bench}, a fine-grained egocentric benchmark for grounded and verifiable operation-centric reasoning with explicit step-by-step rationale annotations. Overall, EgoCoT-Bench comprises 3,172 verifiable QA pairs over 351 egocentric videos separated into four task groups for a total of 12 sub-task groups, encompassing perception and retrospection, anticipation, and high-level reasoning. The benchmark is constructed through a spatio-temporal scene graphs (STSG) guided generation framework and is further refined by human annotators to ensure correctness, egocentric relevance and fine-grained quality. Experimental results show continuing difficulties with egocentric fine-grained reasoning and further reveal that many multimodal models produce explanations that are answer-correct, but have evidence that is inconsistent with the answer. We hope EgoCoT-Bench can serve as a useful testbed for grounded and verifiable reasoning in egocentric video understanding. Project page and supplementary materials are available at: https://dstardust.github.io/EgoCoT/.
arXiv:2605.20147v1 Announce Type: new
Abstract: Text-to-Image (T2I) models have recently seen notable progress around 1K and 2K resolution. With the extreme desire for better visual experience and the rapid development of imaging technology, the demand for Ultra-High-Resolution (UHR) image generation has grown significantly. However, UHR image generation poses great challenges due to the scarcity and complexity of high-resolution content. In this paper, we first introduce PixVerve-95K, a high-quality, open-source UHR T2I dataset curated with a carefully designed data pipeline, which contains 95K images across diverse scenarios (each image has a minimum pixel-count of 100M) and seven-dimensional annotations. Based on our large-scale image-text dataset, we take a pioneering step to extend various T2I foundation models to native 100MP generation with three training schemes. Finally, leveraging both conventional metrics and multimodal large language model-based assessments, our proposed PixVerve-Bench benchmark establishes a comprehensive evaluation protocol for UHR images encompassing visual quality and semantic alignment. Extensive experimental results on our benchmark and the constructive exploration of training strategies collaboratively provide valuable insights for future breakthroughs.
arXiv:2605.19561v1 Announce Type: new
Abstract: As Large Language Models (LLMs) advance toward practical deployment, the Microscaling FP4 (MXFP4) format has emerged as a cornerstone for next-generation low-bit inference, owing to its ability to balance high dynamic range with hardware efficiency. However, directly applying MXFP4 to LLM activation quantization inevitably leads to significant accuracy degradation. In this paper, we theoretically analyze the error structure of MXFP4 activation quantization, revealing that the root cause of this performance drop lies in two structural imbalances between activation distributions and the MXFP4 block floating-point format: (1) extreme inter-block variance imbalance and (2) intra-block codebook utilization imbalance. To address these challenges, we propose TORQ (Two-level Orthogonal Rotation for MXFP4 Quantization), a training-free Post-Training Quantization (PTQ) framework designed to reshape the geometric properties of the activation space through optimal coordinate transformations. At the macroscopic level, TORQ leverages the Schur-Horn theorem to redistribute activation energy via inter-block orthogonal rotation, preventing high-variance blocks from driving up shared scaling factors and thereby preserving the precision of small-magnitude elements. At the microscopic level, TORQ employs maximum-entropy-guided intra-block rotation to alleviate codebook collapse and maximize the MXFP4 codebook's information capacity. Experiments on mainstream LLMs such as LLaMA3 and Qwen3 show that TORQ significantly improves the accuracy of MXFP4 activation quantization compared to existing methods: on Qwen3-32B, the perplexity on WikiText is reduced to 8.43 (vs. 7.61 for BF16), and the average accuracy increases from 38.40% with direct RTN to 73.63% (vs. 74.82% for BF16), substantially narrowing the gap between 4-bit floating-point quantization and full-precision inference.