arXiv:2607.05620v1 Announce Type: new Abstract: In many decision-making settings, new interventions are acceptable only if they do not reduce outcomes below some established threshold. For example, in clinical medicine, new treatments are often acceptable only if they do not worsen outcomes relative to an established standard of care. Safe Bayesian optimization maximizes an objective subject to safety constraints. In the setting that we consider here, safety is defined relative to a known baseline policy whose outcomes are counterfactual and therefore unobserved. Thus, the counterfactual outcomes of the baseline policy must be estimated and those (uncertain) estimates must be used to safely optimize the objective. We address this estimation problem by using conformal prediction to construct valid uncertainty intervals for counterfactual baseline outcomes, and we show how these intervals can be integrated into safe Bayesian optimization to ensure that constraint violations occur at or below a user-specified rate. We also show how to adapt these conformal estimates to different kinds of covariate shift. We provide a safety proof, experimental evidence, and a sensitivity analysis.
Science Journals
arXiv:2607.05815v1 Announce Type: new Abstract: The Minkowski functionals of a field's excursion sets -- area, boundary measure, and Euler characteristic -- describe its level-set morphology; the Euler characteristic is the cheapest handle on topology. We derive smooth Monte-Carlo estimators for all three of a continuous neural field, evaluated at scattered points via the co-area formula and Gauss-Bonnet, using only autodiff: no grid, no complex, no persistence. The estimator is accurate to 1-3% against exact topology in 2D and 3D, and costs about 3 ms per iteration where a persistent-homology (PH) loss on a cubical grid costs 650-1000 ms -- a 250x gap. We establish four design rules without which these losses silently fail: a dense level ladder (invariants are flat in the parameters away from transitions), a $C^2$ backbone (ReLU nets hide curvature in kinks), the full Minkowski vector (Euler characteristic alone is an alternating sum, gamed by debris-hole cancellation; pricing perimeter closes the channel), and sampling-scale coverage. In 2D the vector-valued cap is the only method in a controlled comparison that both repairs topology (3/3 seeds) and preserves fidelity -- uniform smoothing repairs at 11-17x the fidelity cost, and the Euler term alone repairs nothing. In 3D neural-SDF fitting, however, a failure mode we believe general to any sampled soft topology objective appears: gradient descent adversarially hides topological noise below the sampling density, where the estimator is blind -- spurious-feature counts are invariant to 4x more samples, and closing the window needs cubically many points, erasing the cost advantage. A grid-based PH baseline, whose complex is the evaluation resolution, solves the same benchmark ($4/9$ exact; median $b_1$ error 1 vs. ours above $10^4$). The 250x cost of persistence is, at present, the price of having no null space. We release estimators, receipts, and benchmarks.
arXiv:2607.05635v1 Announce Type: new Abstract: Random Vector Functional Link (RVFL) networks are popular due to their fast training and universal approximation capabilities. However, RVFL models face challenges in preserving geometric relationships and utilizing multiple feature views effectively. To address these limitations we propose the Intuitionistic Fuzzy Graph Embedded Random Vector Functional Link with Multiview Learning (IFGRVFL-MV) model. The proposed approach comprises three key components: intuitionistic fuzzy sets for uncertainty handling, graph embedding to capture intrinsic geometric structures, and multiview learning to use complementary information from multiple feature spaces. The model assigns intuitionistic fuzzy membership and non-membership values to data points making it robust to outliers. Also, the graph embedding framework preserves topological structures, increasing the generalization performance. We performed experiments on benchmark datasets from UCI and KEEL repositories which concludes that IFGRVFL-MV outperforms existing models in classification accuracy. Our results establish that IFGRVFL-MV is a promising advancement in the domain of uncertainty and multiview environments.
arXiv:2607.05657v1 Announce Type: new Abstract: This work presents NEMESIS, a multimodal framework for operational transconductance amplifier (OTA) design using large language models (LLMs). NEMESIS strikes a balance between fast, approximate analytical models vs. accurate, computationally expensive SPICE evaluations. Given an OTA netlist and schematic, NEMESIS first identifies circuit primitives and then generates progressively more accurate performance equations. The framework begins with equations retrieved from the prior invocations of NEMESIS to structurally similar OTAs, if available; otherwise, it uses the LLM to derive the initial equations directly from the circuit input. These equations are iteratively refined via a SPICE-based repair loop. In a commercial 65nm PDK, NEMESIS is demonstrated on five OTA topologies, producing SPICE-verified equations across biasing ranges with <7% average relative error and a post-convergence evaluation speedup of ~4622x over full SPICE-based evaluation.
arXiv:2607.05679v1 Announce Type: new Abstract: Language models (LMs) exhibit problematic biases, such as stereotypes. Effectively analyzing and mitigating such biases requires accurate and generalizable evaluation methods of the underlying associations. Some existing approaches focus on downstream metrics that analyze associations in generated text. Since generated text content can vary drastically across LMs, such metrics often require specialized evaluation datasets, which limits the generalization of such downstream metrics. In contrast, upstream metrics examine LMs at the fundamental level of embeddings or continuation probabilities, enabling principled association analyses across LMs. Yet, to date, no upstream metric for generative LMs has uncovered a strong relationship with real-world associations, including those measured in generated text. To address this gap, we introduce the Relative Probability Association Metric (RPAM), an association evaluation metric for generative LMs. For three LMs of different quality of language generation and purpose (Mistral-7B-Instruct, Mistral-7B, and GPT-2) and well-studied evaluation datasets (WEAT-WS, Bellezza, WS-353, and SST2), we find a strong relationship between upstream RPAM measurements and corresponding implicit and explicit associations observed in humans, as well as biases measured downstream with LM-specific tasks, outperforming prior record values where applicable.
arXiv:2607.05684v1 Announce Type: new Abstract: Efficient concentration and transport of electromagnetic energy through electromagnetically thick structures often requires resonant phenomenon and careful design considerations. Here, we introduce a realistic non-resonant approach based on electromagnetically thick self-dual metasurfaces that can funnel electromagnetic waves through subwavelength regions and without requiring magnetic materials. By satisfying the self-duality condition, the proposed metasurfaces support impedance-matched propagation and enable reflectionless energy transfer regardless of the metasurface thickness or structural details. This mechanism also allows selected control over the internal field while maintaining reasonable operational bandwidth. Metasurface elements are designed individually, and full-wave simulations confirm the predicted behavior in sample representative cases. The proposed framework provides a general strategy for robust electromagnetic energy routing and confinement, with potential impact in nonlinear optics, sensing and particle manipulation, near-field imaging, advanced absorber technologies, and wide-angle antenna systems.
arXiv:2607.05702v1 Announce Type: new Abstract: Images acquired in surveillance environments often suffer from conditions such as low resolution, variations in pose, irregular illumination, and occlusions. Due to the low quality of these images, face recognition algorithms often struggle. This major limitation can be addressed by employing super-resolution techniques that enhance the details of the image. However, due to the high degree of difficulty of the problem, most super-resolution algorithms tend to cause distortions in the image and in the individual's identity. Thus, additional information must be incorporated into the processing to improve recognition robustness. In this regard, surveillance cameras can capture multiple images, even at low quality, and the data extracted from these images, such as consecutive video frames, can significantly enhance both super-resolution and facial recognition. In this work, we introduce FASR++, a diffusion-model-based super-resolution algorithm. It leverages a reference low-resolution image and features extracted from multiple auxiliary low-quality images to generate a super-resolved output, minimizing distortions in the individual's identity. Our approach recovers facial features without explicitly providing soft attributes or computing a function gradient to guide the reconstruction process. FASR++ generates high-quality images that can considerably improve performance in face recognition tasks when used as a pre-processing step. We validate our approach on two standard face recognition datasets and attain state-of-the-art results for verification, face recognition, and image quality metrics such as PSNR, SSIM, and LPIPS.
arXiv:2607.05708v1 Announce Type: new Abstract: Recent LLM-based agent systems continuously accumulate context across multi-turn interactions, tool invocations, and cross-session workflows. Replaying the full history for every request quickly becomes impractical: long contexts increase prefill cost, may exceed context limits, and often bury task-relevant evidence in irrelevant content, degrading both serving efficiency and output quality. We propose Akashic, a low-overhead memory system built around MemAttention, which organizes context into bounded chunks and models semantic relationships across chunks, preserving cross-chunk evidence without repeatedly rewriting the full history. Akashic further applies hardware-software co-designed memory placement to co-locate likely co-retrieved chunks, reducing retrieval fragmentation and I/O overhead. Across four representative workloads and three model sizes, Akashic improves task accuracy by up to 10.2 points, throughput by up to 1.21x, and sustainable request rate by up to 1.88x over strong prior memory baselines.
arXiv:2607.05785v1 Announce Type: new Abstract: Recent advances in coding agents have enabled the generation of increasingly complex software systems. While existing evaluations primarily focus on functional correctness, production systems must expose failure evidence to support observability. In this paper, we present a systematic study of observability in agent-generated systems. We examine whether agents can reconstruct source-level diagnostic semantics by restoring observability artifacts in 10 open-source and 8 industrial repositories. We also evaluate whether these artifacts translate into effective fault signals at runtime through 200 generated microservice systems deployed on Kubernetes with 13 injected faults. Our results reveal a consistent gap between diagnostic semantics at the source level and fault signals (i.e., explicit, fault-specific evidence) at runtime. At the source level, agents partially recover observability artifacts but struggle to capture key diagnostic semantics. At runtime, generated systems expose fault signals for only a small fraction of failures (up to 13.99\%), despite the presence of logging, suggesting that the generated observability artifacts may lack the failure-specific semantics needed to effectively expose faults. We further introduce an observability-oriented skill, which can serve as a guidance to improve both diagnostic semantics and fault-signal exposure, but the gains remain limited, indicating that the gap is not easily addressed. More broadly, our findings suggest that current evaluations focusing primarily on functional correctness may overlook observability as an important dimension of practical software quality.
arXiv:2607.05811v1 Announce Type: new Abstract: We consider the numerical solution of the wave equation in materials with rapidly varying coefficients, and time harmonic sources. For these problems, direct discretization is prohibitively costly, and instead multiscale methods are used. There are several multiscale methods that directly discretize in the frequency domain. In this work we instead start in the time-domain and combine a finite difference Heterogeneous Multiscale Method (HMM) for the wave equation with the WaveHoltz method. Each WaveHoltz iteration marches the wave equation towards the time-periodic Helmholtz solution. The advantages of the WaveHoltz method relative to traditional Helmholtz solvers carry over directly to the multiscale problems considered here. Since, in addition, the time-domain solver does not artificially impose boundary conditions on the micro-scale problems, no boundary errors from the micro-scale problems are present in the homogenized frequency domain solution.
arXiv:2607.05820v1 Announce Type: new Abstract: Large-scale circular gap closure occurs over a time scale on which cell growth and proliferation become important. Growth is the main driver of the closing process, while cell dynamics such as elongation and intercalation reflect elastic and fluidic contributions to tissue deformation. We develop a novel fluidized growth-elasticity framework as a nonlinear analogue of a Maxwell fluid with growth. The framework decomposes the experimentally observable strain rate into the additive sum of the growth, elastic, and fluidic strain rates, thus enabling the separate quantification of these contributions from tissue kinematics and allowing the roles of tissue elasticity and fluidity (the inverse of viscosity) to be characterized. We apply the model to large circular gaps ($\sim$1.7 mm in diameter) in confluent monolayers of mouse embryonic epicardial cells (MEC1) under two conditions, without and with TGF-$\beta$ treatment. We show that both tissue fluidity and the elastic properties associated with fiber reinforcement are critical for reproducing the closure kinematics. Specifically, we predict that the treated condition has lower fluidity, associated with a lower fluidic deformation rate and a higher elastic deformation rate than the untreated condition, in agreement with the experimental observations.
arXiv:2607.05859v1 Announce Type: new Abstract: Vision-Language Models (VLMs) are promising for construction-site monitoring, and recent construction-tailored VLMs have primarily adapted pretrained VLMs through direct QA-style fine-tuning from a single global image. We argue that this direct paradigm remains limited for in-the-wild deployment in terms of operational range, reliability under reduced-resolution inputs, and inference efficiency. To address these challenges, we propose AVA-VLM, an Adaptive Visual Attention-Vision Language Model that follows a human-inspired coarse-to-fine reasoning strategy. AVA-VLM first reasons over a low-resolution global image and selectively requests a high-resolution local crop only when detailed inspection is needed, similar to how a human inspector zooms in on hard-to-see yet important areas. We further introduce a region-aware Chain-of-Thought dataset that teaches the model when to inspect, where to crop, and how to use local evidence. Experiments show that AVA-VLM improves reliability under long-distance and reduced-resolution conditions while substantially reducing visual-token usage.
arXiv:2607.06292v1 Announce Type: new Abstract: The use of Gaussian processes for approximating differential equations has expanded rapidly, leading to a growing, diverse, and fragmented body of numerical methods. We present a unified Bayesian perspective that places these techniques within a common probabilistic framework, based on a derivative matching interpretation for incorporating differential equation constraints into likelihood. This unified perspective supports both parameter estimation and solution approximation, and shows how a range of existing methods can be understood within it. This work aims to consolidate current developments and provide a foundation for future research.
arXiv:2607.05824v1 Announce Type: new Abstract: High-throughput holotomography often relies on long-working-distance, multiwell-compatible optics that reduce illumination numerical aperture (NA) and limit access to high spatial frequencies. Here we present ResShift-ODE, a deterministic diffusion-prior framework that transfers low-NA refractive-index (RI) tomograms to high-NA-equivalent volumes without modifying the acquisition hardware. We formulate low-NA-to-high-NA transfer as a diffusion-prior inverse problem under an explicit NA-limited Fourier-domain forward operator, distinct from suppressing artifacts within an already measured passband. The method extends residual-shifting diffusion to volumetric RI data and reformulates the reverse process as a probability-flow ordinary differential equation, enabling reproducible inference in five denoiser evaluations. On held-out emulated-pair test volumes, inferred volumes matched high-NA references with RI errors of 0.002-0.003 for >99% of voxels, while Fourier analysis confirmed measurement-anchored lateral-band recovery without filling the axial missing cone. 3D ResShift-ODE required five denoiser evaluations per volume, incurring ~5.8x the cost of a 3D U-Net while remaining ~166x faster than a 1000-step 3D denoising diffusion probabilistic model.
arXiv:2607.05837v1 Announce Type: new Abstract: Defocus blur degrades fine image structures and limits visual perception, which can adversely affect downstream vision tasks. Although recent deep learning deblurring methods have achieved strong performance, their effectiveness depends on training data and often degrades across cameras and lenses due to limited optical diversity and realism in existing datasets. In this paper, we propose a pipeline for synthesizing realistic defocus deblurring datasets for diverse compound lenses. It integrates efficient wave-optics PSF computation via Debye CZT propagation, depth-aware defocus rendering with occlusion handling, and blur synthesis in the radiometrically linear space with camera ISP simulation. This unified pipeline enables the scalable generation of photorealistic defocus datasets with diverse lens characteristics. Using our pipeline, we generate CLDefocus, a large-scale synthetic dataset containing lens-diverse defocus image pairs. We further analyze the limitations of real-captured defocus datasets and show that such imperfections can bias full-reference evaluation. Extensive experiments demonstrate that models trained on CLDefocus achieve improved cross-device generalization compared to models trained on existing real and synthetic datasets.
arXiv:2607.05846v1 Announce Type: new Abstract: Accurate ranking of antibody candidates according to their binding affinity is essential for therapeutic antibody discovery. However, existing methods treat affinity comparisons independently and ignore the contextual information encoded in other labeled comparisons, limiting their ability to capture antigen-specific binding landscapes. For many target antigens, a small number of experimentally characterized affinity comparisons are often available. An important question is whether the model can exploit these existing comparisons to infer antigen-specific ranking patterns that facilitate subsequent affinity ranking. This form of learning from labeled demonstrations closely resembles the paradigm of In-Context Learning, motivating us to revisit antibody affinity ranking from an ICL perspective. To this end, we propose AbICL, an ICL framework for antigen-specific antibody affinity ranking. AbICL combines a pretrained structural encoder with a context ranking head and is trained with an episodic meta-training strategy that enables the model to leverage support demonstrations for test-time adaptation without gradient updates. Experiments on the AbRank benchmark demonstrate that AbICL consistently outperforms existing ranking baselines across almost all data splits and evaluation benchmarks. Further analysis shows that the value of contextual demonstrations depends on how well they match the target inference task, and becomes increasingly pronounced under distribution shift and fine-grained affinity discrimination. These findings highlight the potential of ICL as an effective paradigm for antigen-specific antibody affinity ranking, particularly in challenging settings where a single global ranking function is insufficient.
arXiv:2607.05850v1 Announce Type: new Abstract: Deep neural networks trained with Empirical Risk Minimization (ERM) often fail under distribution shifts because they exploit spurious correlations between object labels and background context. Recent generative approaches address this issue by creating counterfactual images with altered contexts, but typically use these samples as standard data augmentation, leaving the model free to retain background-sensitive representations. We propose a two-stage framework that uses generative intervention to explicitly learn background-invariant visual representations. First, we isolate the foreground object using zero-shot segmentation and generate context-shifted variants with a structure-preserving diffusion model, preserving object identity while varying the surrounding environment. We then introduce Cross-Variant Self-Supervised Learning, where variants of the same object under different backgrounds form positive pairs in a contrastive objective. This encourages the encoder to align object-centric representations while suppressing background-specific cues. Then, we fine-tune the pretrained encoder using an ERM warm-up followed by GroupDRO with layer-wise learning rates. Experiments on distribution-shift benchmarks demonstrate best worst-group performance, achieving 92.5% on Waterbirds, 81.7% on MetaShift, and 87.4% on NICO++. Code: https://github.com/surajyadav-research/GRSSL
arXiv:2607.05855v1 Announce Type: new Abstract: Information Operations on social media networks have been identified as a significant threat to democracy and modern society, but they are challenging and expensive to detect by humans. Existing supervised IO detection methods fail to capture the dynamic nature of evolving IO user behavior, while existing unsupervised approaches rely on oversimplified assumptions of coordination among IO users that may not exist in practice. To overcome the limitations of existing methods, we formulate IO user detection as an anomaly detection problem and propose a novel unsupervised IO user detection approach called Temporal-bEhavior-laNguage Signals for information Operation Recognition (TENSOR), which leverages multimodal data, including temporal online user behavior, such as message posting activities, and the textual content of the messages. The motivation is that IO users are typically a very small fraction of all online users and have unique temporal behavioral and language patterns. Specifically, we train a Temporal Point Process (TPP) to capture abnormal temporal behavioral patterns of IO users because they are known to behave in a coordinated manner for IO campaigns. We further introduce a novel evidence function that converts LLM responses, which are generated from user post timelines, into quantitative scores to adjust the TPP outputs for better IO user detection. Experimental results show that TENSOR outperforms the baselines on five real-world IO datasets. Code is available at https://github.com/xiuzhenzhang/TENSOR.
arXiv:2607.05867v1 Announce Type: new Abstract: Dynamic shaping of ultrasound into prescribed spatial patterns underlies a broad range of biomedical and engineering applications. However, existing modulation strategies face fundamental limitations: single element transducers paired with acoustic lenses lack reconfigurability, whereas phased arrays require large numbers of independently driven elements, leading to substantial hardware complexity, cost, and rigidity. Here we introduce a microfluidic ultrasound lens system that enables reconfigurable spatial modulation of ultrasonic fields using two orthogonal layers of soft microfluidic channels. Each channel is selectively filled with one of two liquids with distinct sound speeds via an FPGA controlled array of micropumps, generating programmable binary phase patterns. Integrating a 20-row-by-20-column microfluidic lens with a single element transducer, we demonstrate three-dimensional ultrasound focusing with approximately one second reconfiguration time and spatial resolution comparable to that of a 400-element transducer array. The system provides 400 addressable pixels through parallel control of 80 pumps, allowing hardware complexity to scale with the square root of the pixel count. Building on this platform, we demonstrate dynamic ultrasound heating, as well as remote particle manipulation. Furthermore, we demonstrate a cylindrical lens that manipulates ultrasound propagation in the azimuthal direction. Owing to its liquid based, soft architecture, the microfluidic lens offers design flexibility, scalable operation across ultrasound frequencies, low acoustic transmission loss, and stable performance under high acoustic power. Together, these results establish microfluidic phase modulation as a compact, scalable, and flexible approach for dynamic ultrasound field control.
arXiv:2607.05869v1 Announce Type: new Abstract: Robust robotic grasping of novel objects requires datasets that simultaneously provide photorealistic RGB-D observations, physically validated grasp quality annotations, and a principled bridge between simulation and the real world, which existing datasets lack to provide jointly. \textbf{GraspIT} addresses this gap: tabletop scenes in NVIDIA Isaac Sim are annotated via a four-stage physical slip-test on parallel Franka Panda instances, producing trajectory-reachability checks and continuous quality scores beyond force-closure.Of ${\sim}$2.3M candidates, 83% pass as \emph{good} ($s{\geq}0.50$); the 17% that passed force-closure but failed the slip-test provide graded hard negatives. A Real$\leftrightarrow$Sim loop back-projects these labels onto 100 real-world scenes. The release provides ${\sim}$316k annotated RGBD frame sets across 1035 sim and 100 real scenes, with instance masks, 6-DoF poses, physical object properties, and scored 6-DoF grasps. All tools are open-source and Docker-containerized. The trajectory planning within Isaac Sim further allows streaming of high resolution demonstrations for tabletop manipulation policy learning and behavior cloning.
arXiv:2607.05877v1 Announce Type: new Abstract: The concept of Digital Twin (DT) consists of a physical asset, a digital asset, and their bidirectional data exchange, differing the DT from concepts with lower level of integration. Availability of the bidirectional interconnection not only enables monitoring of system states but also allows automated control of the system. Leveraging the DT concept, this paper presents a practical implementation of an electric drive DT by means of a permanent magnet synchronous machine (PMSM) in a test lab environment. The approach aims to virtually sense the disturbance variable, here the acting load torque. By integrating an observer controller into the PMSM control loop, the DT enables compensation of the disturbance variable, which is not measurable in many applications. The results of the DT implementation recorded on the test bench demonstrate an effective disturbance compensation of the acting load torque by the observer controller. However, the results still show a deviation between the observed and measured load torques, indicating room for further refinement. This work serves as a first step towards further practical applications of the DT in the electric drive testing environment.
arXiv:2607.05883v1 Announce Type: new Abstract: In dual-arm dexterous teleoperation, cross-platform generalization of motion retargeting and interactivity of grasping are crucial. However, the heterogeneity of robotic architectures and the wide variety of grasping objects pose significant challenges to achieving precise motion retargeting and compliant grasping in dual-arm dexterous teleoperation. To address these challenges, a dual-arm dexterous teleoperation system (DexTele) is proposed based on motion retargeting and adaptive force control. First, a vision-based motion retargeting module is designed to generate preliminary robot motions from human images. In this module, a motion-graph encoder and latent optimization are proposed for precise and convenient cross-platform motion retargeting. Second, an adaptive grasping module is designed to achieve compliant grasping. This module combines a vision-language model (VLM) with model predictive control (MPC), allowing the system to predict the required grasping force for a target object and perform gradient-based online optimization. Finally, extensive experiments demonstrate that the DexTele achieves precise motion retargeting and compliant grasping with generalization across multiple robot platforms.
arXiv:2607.05889v1 Announce Type: new Abstract: Connected Automated Vehicles (CAVs) utilize their onboard sensors to perceive the environment. The perception range and accuracy can be affected by adverse weather or non-line-of-sight conditions. Cooperative perception or sensor sharing can overcome these limitations by enabling CAVs to exchange sensor data, thus collectively enhancing their perception capabilities. Previous studies have shown the potential of cooperative perception, but limited attention has been given to the fusion of V2X data received through cooperative perception messages with onboard sensor information. The fusion process can be influenced by the quantity and quality of the V2X data. An increased volume of V2X data can reduce uncertainty in the perceived environment; however, when the data is noisy, it may compromise the accuracy of the fusion results. This study investigates the fusion of onboard sensor and V2X data in cooperative perception, and demonstrates that while perception can significantly improve as the V2X penetration rate increases, it can introduce a significant number of false positives if V2X data is not highly accurate. False positives result in the detection of ghost objects that do not actually exist. These ghost objects can, in turn, compromise safety and driving efficiency. Our analysis found that false positives or ghost objects can appear even with accurate V2X data. These findings highlight the challenges in cooperative perception and the importance of developing robust data fusion methods to enhance the reliability of cooperative perception. This is particularly relevant in light of ongoing standardization efforts, such as ETSI TS 103 324 on collective perception.
arXiv:2607.06440v1 Announce Type: new Abstract: Recent text-to-image models such as DALLE-3 excel at following diverse prompts yet remain blind to individual aesthetic preferences. We study personalized image generation, where models must align outputs with a user's implicit visual preferences based on a few historically preferred images and a short prompt. To this end, we introduce PIPBench, the first profile-inclusive benchmark for evaluating personalized image generation. We further propose a novel data construction pipeline that leverages psychological and demographic profiling dimensions for both real-user data collection and scalable agent-based data generation. Using PIPBench, we conduct a thorough evaluation of representative line of methods. Our experiments reveal key limitations in existing methods, suggesting new challenges and opportunities for personalized text-to-image synthesis. Project page: https://wuyuhang05.github.io/PIPBench/
arXiv:2607.05911v1 Announce Type: new Abstract: Compositional Zero-Shot Learning (CZSL) aims to combine known attributes and objects as primitives for recognizing previously unseen attribute-object pairs. Prior works either predict attributes and objects independently, missing their strong contextual dependency, or use unidirectional conditional modeling (e.g., object-guided attribute prediction), which is prone to error propagation. We propose PRPC, a Progressive Reasoning framework with Primitive Correction, which explicitly models the bidirectional dependency between attributes and objects via step-wise inference. PRPC performs mutual correction of primitives to suppress prediction errors in earlier steps. Specifically, we formulate CZSL as structured, Q&A-style Chain-of-Thought reasoning process and constrain the MLLM to follow predefined semantic steps to generate intermediate decisions. To further enhance the reliability and logical consistency of intermediate reasoning, we introduce reinforcement learning post-training with a GRPO-based objective, providing step-level rewards aligned with the progressive inference procedure. Extensive experiments on three CZSL benchmarks demonstrate that PRPC achieves state-of-the-art performance, validating the effectiveness of progressive reasoning and bidirectional correction for robust compositional generalization.