RADAR: Redundancy-Aware Diffusion for Multi-Agent Communication Structure Generation

Compared with individual agents, large language model based multi-agent systems have shown great capabilities consistently across diverse tasks, including code generation, mathematical reasoning, and planning, etc. Despite their impressive performance, the effectiveness and robustness of these systems heavily rely on their communication topology, which is often fixed or generated in a single step. This restricts fine-grained structural exploration and flexible composition, resulting in excessive token utilization on simple tasks while limiting capability on complicated tasks. To mitigate this challenge, we introduce RADAR, a redundancy-aware and query-adaptive generative framework that actively reduce communication overhead. Motivated by recent progress in conditional discrete graph diffusion models, we formulate communication topology design as a step-by-step generation process, guided by the effective size of the graph. Comprehensive experiments on six benchmarks demonstrate that RADAR consistently outperforms recent baselines, achieving higher accuracy, lower token consumption, and greater robustness across diverse scenarios. Our code and data are available at https://github.com/cszhangzhen/RADAR.

AffectSeek: Agentic Affective Understanding in Long Videos under Vague User Queries

Existing affective understanding studies have mainly focused on recognizing emotions from images, audio signals, or pre-cliped video clips, where the affective evidence is already given. This passive and clip-centered setting does not fully reflect real-world scenarios, in which users often interact with long videos and express their needs through natural-language queries. In this paper, we study \textbf{Vague-Query-driven video Affective Understanding (VQAU)}, a new task that requires models to localize affective moments in long videos, predict their emotion categories, and generate evidence-grounded rationales under vague user queries. To support this task, we construct \textbf{VQAU-Bench}, a benchmark that integrates long videos, vague affective queries, temporal clip annotations, emotion labels, and rationale explanations into a unified evaluation framework. VQAU-Bench enables systematic assessment of semantic-temporal-affective alignment, affective moment localization, emotion classification, and rationale generation. To address the multi-step reasoning challenges of VQAU, we further propose \textbf{AffectSeek}, an agentic framework that actively seeks, verifies, and explains affective moments in long videos. AffectSeek decomposes VQAU into intent interpretation, candidate localization, clip verification, emotion reasoning, and rationale generation, and progressively aligns vague user intent with long-video evidence through role-specialized reasoning and cross-stage verification. Experiments show that VQAU remains challenging for existing affective recognition models and single-step vision-language models, while AffectSeek provides a simple yet effective framework for agentic long-video affective understanding.

Causal Disentanglement for Full-Reference Image Quality Assessment

Existing deep network-based full-reference image quality assessment (FR-IQA) models typically work by performing pairwise comparisons of deep features from the reference and distorted images. In this paper, we approach this problem from a different perspective and propose a novel FR-IQA paradigm based on causal inference and decoupled representation learning. Unlike typical feature comparison-based FR-IQA models, our approach formulates degradation estimation as a causal disentanglement process guided by intervention on latent representations. We first decouple degradation and content representations by exploiting the content invariance between the reference and distorted images. Second, inspired by the human visual masking effect, we design a masking module to model the causal relationship between image content and degradation features, thereby extracting content-influenced degradation features from distorted images. Finally, quality scores are predicted from these degradation features using either supervised regression or label-free dimensionality reduction. Extensive experiments demonstrate that our method achieves highly competitive performance on standard IQA benchmarks across fully supervised, few-label, and label-free settings. Furthermore, we evaluate the approach on diverse non-standard natural image domains with scarce data, including underwater, radiographic, medical, neutron, and screen-content images. Benefiting from its ability to perform scenario-specific training and prediction without labeled IQA data, our method exhibits superior cross-domain generalization compared to existing training-free FR-IQA models.

FingerViP: Learning Real-World Dexterous Manipulation with Fingertip Visual Perception

The current practice of dexterous manipulation generally relies on a single wrist-mounted view, which is often occluded and limits performance on tasks requiring multi-view perception. In this work, we present FingerViP, a learning system that utilizes a visuomotor policy with fingertip visual perception for dexterous manipulation. Specifically, we design a vision-enhanced fingertip module with an embedded miniature camera and install the modules on each finger of a multi-fingered hand. The fingertip cameras substantially improve visual perception by providing comprehensive, multi-view feedback of both the hand and its surrounding environment. Building on the integrated fingertip modules, we develop a diffusion-based whole-body visuomotor policy conditioned on a third-view camera and multi-view fingertip vision, which effectively learns complex manipulation skills directly from human demonstrations. To improve view-proprioception alignment and contact awareness, each fingertip visual feature is augmented with its corresponding camera pose encoding and per-finger joint-current encoding. We validate the effectiveness of the multi-view fingertip vision and demonstrate the robustness and adaptability of FingerViP on various challenging real-world tasks, including pressing buttons inside a confined box, retrieving sticks from an unstable support, retrieving objects behind an occluding curtain, and performing long-horizon cabinet opening and object retrieval, achieving an overall success rate of 80.8%. All hardware designs and code will be fully open-sourced.

Paradigm Shift from Statistical Channel Modeling to Digital Twin Prediction: An Environment-Generalizable ChannelLM for 6G AI-enabled Air Interface

As 6G advances, ubiquitous connectivity and higher capacity requirements of the air interface pose substantial challenges for accurate and real-time wireless channel acquisition in diverse environments. Conventional statistical channel modeling relies on offline measurement data from limited environments, struggling to support online applications facing diverse environments. To this end, the digital twin channel (DTC) has emerged as a novel paradigm that constructs a digital replica of the physical environment through high-fidelity sensing and predicts corresponding channel in real time utilizing artificial intelligence (AI) models. As the engine of DTC, existing AI models struggle to simultaneously achieve strong environmental generalization in real-world and end-to-end channel prediction for real time tasks. Therefore, this paper proposes a channel large model (ChannelLM)-driven DTC architecture comprising three modules: low-complexity and high-accuracy environment reconstruction based on dynamic object detection and multimodal alignment of image and point cloud data, physically interpretable environment feature extraction, and a ChannelLM core to mapping these features into generalized environment representations for multi-task channel prediction. Simulation results demonstrate that, in unseen test environments, compared with small-scale AI models, ChannelLM reduces prediction errors by 4.23 dB in channel state information prediction while achieving an end-to-end inference latency of 70 milliseconds in the real world.

PINNs in PDE Constrained Optimal Control Problems: Direct vs Indirect Methods

We study physics-informed neural networks (PINNs) as numerical tools for the optimal control of semilinear partial differential equations. We first recall the classical direct and indirect viewpoints for optimal control of PDEs, and then present two PINN formulations: a direct formulation based on minimizing the objective under the state constraint, and an indirect formulation based on the first-order optimality system. For a class of semilinear parabolic equations, we derive the state equation, the adjoint equation, and the stationarity condition in a form consistent with continuous-time Pontryagin-type optimality conditions. We then specialize the framework to an Allen-Cahn control problem and compare three numerical approaches: (i) a discretize-then-optimize adjoint method, (ii) a direct PINN, and (iii) an indirect PINN. Numerical results show that the PINN parameterization has an implicit regularizing effect, in the sense that it tends to produce smoother control profiles. They also indicate that the indirect PINN more faithfully preserves the PDE contraint and optimality structure and yields a more accurate neural approximation than the direct PINN.

When Users Change Their Mind: Evaluating Interruptible Agents in Long-Horizon Web Navigation

As LLM agents transition from short, static problem solving to executing complex, long-horizon tasks in dynamic environments, the ability to handle user interruptions, such as adding requirement or revising goals, during mid-task execution is becoming a core requirement for realistic deployment. However, existing benchmarks largely assume uninterrupted agent behavior or study interruptions only in short, unconstrained language tasks. In this paper, we present the first systematic study of interruptible agents in long-horizon, environmentally grounded web navigation tasks, where actions induce persistent state changes. We formalize three realistic interruption types, including addition, revision, and retraction, and introduce InterruptBench, a benchmark derived from WebArena-Lite that synthesizes high-quality interruption scenarios under strict semantic constraints. Using a unified interruption simulation framework, we evaluate six strong LLM backbones across single- and multi-turn interruption settings, analyzing both their effectiveness in adapting to updated intents and their efficiency in recovering from mid-task changes. Our results show that handling user interruptions effectively and efficiently during long-horizon agentic tasks remains challenging for powerful large-scale LLMs. Code and dataset are available at https://github.com/HenryPengZou/InterruptBench.

MathGen: Revealing the Illusion of Mathematical Competence through Text-to-Image Generation

Modern generative models have demonstrated the ability to solve challenging mathematical problems. In many real-world settings, however, mathematical solutions must be expressed visually through diagrams, plots, geometric constructions, and structured symbolic layouts, where correctness depends on precise visual composition. This naturally raises the question of whether generative models can still do so when the answer must be rendered visually rather than written in text? To study this problem, we introduce MathGen, a rigorous benchmark of 900 problems spanning seven core domains, each paired with an executable verifier under a Script-as-a-Judge protocol for deterministic and objective evaluation. Experiments on representative open-source and proprietary text-to-image models show that mathematical fidelity remains a major bottleneck: even the best closed-source model reaches only 42.0% overall accuracy, while open-source models achieve just ~ 1-11%, often near 0% on structured tasks. Overall, current T2I models remain far from competent at even elementary mathematical visual generation.

Cold-Starts in Generative Recommendation: A Reproducibility Study

Cold-start recommendation remains a central challenge in dynamic, open-world platforms, requiring models to recommend for newly registered users (user cold-start) and to recommend newly introduced items to existing users (item cold-start) under sparse or missing interaction signals. Recent generative recommenders built on pre-trained language models (PLMs) are often expected to mitigate cold-start by using item semantic information (e.g., titles and descriptions) and test-time conditioning on limited user context. However, cold-start is rarely treated as a primary evaluation setting in existing studies, and reported gains are difficult to interpret because key design choices, such as model scale, identifier design, and training strategy, are frequently changed together. In this work, we present a systematic reproducibility study of generative recommendation under a unified suite of cold-start protocols.

MiroEval: Benchmarking Multimodal Deep Research Agents in Process and Outcome

Recent progress in deep research systems has been impressive, but evaluation still lags behind real user needs. Existing benchmarks predominantly assess final reports using fixed rubrics, failing to evaluate the underlying research process. Most also offer limited multimodal coverage, rely on synthetic tasks that do not reflect real-world query complexity, and cannot be refreshed as knowledge evolves. To address these gaps, we introduce MiroEval, a benchmark and evaluation framework for deep research systems. The benchmark comprises 100 tasks (70 text-only, 30 multimodal), all grounded in real user needs and constructed via a dual-path pipeline that supports periodic updates, enabling a live and evolving setting. The proposed evaluation suite assesses deep research systems along three complementary dimensions: adaptive synthesis quality evaluation with task-specific rubrics, agentic factuality verification via active retrieval and reasoning over both web sources and multimodal attachments, and process-centric evaluation audits how the system searches, reasons, and refines throughout its investigation. Evaluation across 13 systems yields three principal findings: the three evaluation dimensions capture complementary aspects of system capability, with each revealing distinct strengths and weaknesses across systems; process quality serves as a reliable predictor of overall outcome while revealing weaknesses invisible to output-level metrics; and multimodal tasks pose substantially greater challenges, with most systems declining by 3 to 10 points. The MiroThinker series achieves the most balanced performance, with MiroThinker-H1 ranking the highest overall in both settings. Human verification and robustness results confirm the reliability of the benchmark and evaluation framework. MiroEval provides a holistic diagnostic tool for the next generation of deep research agents.