EQA-RM: A Generative Embodied Reward Model with Test-time Scaling

Reward Models (RMs), vital for large model alignment, are underexplored for complex embodied tasks like Embodied Question Answering (EQA) where nuanced evaluation of agents' spatial, temporal, and logical understanding is critical yet not considered by generic approaches. We introduce EQA-RM, a novel generative multimodal reward model specifically architected for EQA, trained via our innovative Contrastive Group Relative Policy Optimization (C-GRPO) strategy to learn fine-grained behavioral distinctions. The generative nature of EQA-RM provides interpretable, structured reward feedback (beyond simple scalars), uniquely enabling test-time scaling to dynamically adjust evaluation granularity, from concise scores to detailed critiques of reasoning and grounding, at inference without retraining. Concurrently, we introduce EQARewardBench, a new benchmark built on OpenEQA for standardized EQA reward model assessment. Demonstrating high sample efficiency, EQA-RM (fine-tuning Qwen2-VL-2B-Instruct) achieves 61.9\% accuracy on EQA-RM-Bench with only 700 samples, outperforming strong proprietary baselines, including Gemini-2.5-Flash, GPT-4o, Claude-3.5-Haiku, and open-sourced state-of-the-art models such as RoVRM and VisualPRM. The code and dataset can be found here https://github.com/UNITES-Lab/EQA-RM.

IndustryEQA: Pushing the Frontiers of Embodied Question Answering in Industrial Scenarios

Existing Embodied Question Answering (EQA) benchmarks primarily focus on household environments, often overlooking safety-critical aspects and reasoning processes pertinent to industrial settings. This drawback limits the evaluation of agent readiness for real-world industrial applications. To bridge this, we introduce IndustryEQA, the first benchmark dedicated to evaluating embodied agent capabilities within safety-critical warehouse scenarios. Built upon the NVIDIA Isaac Sim platform, IndustryEQA provides high-fidelity episodic memory videos featuring diverse industrial assets, dynamic human agents, and carefully designed hazardous situations inspired by real-world safety guidelines. The benchmark includes rich annotations covering six categories: equipment safety, human safety, object recognition, attribute recognition, temporal understanding, and spatial understanding. Besides, it also provides extra reasoning evaluation based on these categories. Specifically, it comprises 971 question-answer pairs generated from small warehouse and 373 pairs from large ones, incorporating scenarios with and without human. We further propose a comprehensive evaluation framework, including various baseline models, to assess their general perception and reasoning abilities in industrial environments. IndustryEQA aims to steer EQA research towards developing more robust, safety-aware, and practically applicable embodied agents for complex industrial environments. Benchmark and codes are available.

Can Large Multimodal Models Understand Agricultural Scenes? Benchmarking with AgroMind

Large Multimodal Models (LMMs) has demonstrated capabilities across various domains, but comprehensive benchmarks for agricultural remote sensing (RS) remain scarce. Existing benchmarks designed for agricultural RS scenarios exhibit notable limitations, primarily in terms of insufficient scene diversity in the dataset and oversimplified task design. To bridge this gap, we introduce AgroMind, a comprehensive agricultural remote sensing benchmark covering four task dimensions: spatial perception, object understanding, scene understanding, and scene reasoning, with a total of 13 task types, ranging from crop identification and health monitoring to environmental analysis. We curate a high-quality evaluation set by integrating eight public datasets and one private farmland plot dataset, containing 25,026 QA pairs and 15,556 images. The pipeline begins with multi-source data preprocessing, including collection, format standardization, and annotation refinement. We then generate a diverse set of agriculturally relevant questions through the systematic definition of tasks. Finally, we employ LMMs for inference, generating responses, and performing detailed examinations. We evaluated 18 open-source LMMs and 3 closed-source models on AgroMind. Experiments reveal significant performance gaps, particularly in spatial reasoning and fine-grained recognition, it is notable that human performance lags behind several leading LMMs. By establishing a standardized evaluation framework for agricultural RS, AgroMind reveals the limitations of LMMs in domain knowledge and highlights critical challenges for future work. Data and code can be accessed at https://rssysu.github.io/AgroMind/.

Towards Artificial Intelligence Research Assistant for Expert-Involved Learning

Large Language Models (LLMs) and Large Multi-Modal Models (LMMs) have emerged as transformative tools in scientific research, yet their reliability and specific contributions to biomedical applications remain insufficiently characterized. In this study, we present \textbf{AR}tificial \textbf{I}ntelligence research assistant for \textbf{E}xpert-involved \textbf{L}earning (ARIEL), a multimodal dataset designed to benchmark and enhance two critical capabilities of LLMs and LMMs in biomedical research: summarizing extensive scientific texts and interpreting complex biomedical figures. To facilitate rigorous assessment, we create two open-source sets comprising biomedical articles and figures with designed questions. We systematically benchmark both open- and closed-source foundation models, incorporating expert-driven human evaluations conducted by doctoral-level experts. Furthermore, we improve model performance through targeted prompt engineering and fine-tuning strategies for summarizing research papers, and apply test-time computational scaling to enhance the reasoning capabilities of LMMs, achieving superior accuracy compared to human-expert corrections. We also explore the potential of using LMM Agents to generate scientific hypotheses from diverse multimodal inputs. Overall, our results delineate clear strengths and highlight significant limitations of current foundation models, providing actionable insights and guiding future advancements in deploying large-scale language and multi-modal models within biomedical research.

Cross Far- and Near-Field Beam Management Technologies in Millimeter-Wave and Terahertz MIMO Systems

The evolution of wireless communication toward next-generation networks introduces unprecedented demands on data rates, latency, and connectivity. To meet these requirements, two key trends have emerged: the use of higher communication frequencies to provide broader bandwidth, and the deployment of massive multiple-input multiple-output systems with large antenna arrays to compensate for propagation losses and enhance spatial multiplexing. These advancements significantly extend the Rayleigh distance, enabling near-field (NF) propagation alongside the traditional far-field (FF) regime. As user communication distances dynamically span both FF and NF regions, cross-field (CF) communication has also emerged as a practical consideration. Beam management (BM)-including beam scanning, channel state information estimation, beamforming, and beam tracking-plays a central role in maintaining reliable directional communications. While most existing BM techniques are developed for FF channels, recent works begin to address the unique characteristics of NF and CF regimes. This survey presents a comprehensive review of BM techniques from the perspective of propagation fields. We begin by building the basic through analyzing the modeling of FF, NF, and CF channels, along with the associated beam patterns for alignment. Then, we categorize BM techniques by methodologies, and discuss their operational differences across propagation regimes, highlighting how field-dependent channel characteristics influence design tradeoffs and implementation complexity. In addition, for each BM method, we identify open challenges and future research directions, including extending FF methods to NF or CF scenarios, developing unified BM strategies for field-agnostic deployment, and designing low-overhead BM solutions for dynamic environments.

Kinematics-based 3D Human-Object Interaction Reconstruction from Single View

Reconstructing 3D human-object interaction (HOI) from single-view RGB images is challenging due to the absence of depth information and potential occlusions. Existing methods simply predict the body poses merely rely on network training on some indoor datasets, which cannot guarantee the rationality of the results if some body parts are invisible due to occlusions that appear easily. Inspired by the end-effector localization task in robotics, we propose a kinematics-based method that can drive the joints of human body to the human-object contact regions accurately. After an improved forward kinematics algorithm is proposed, the Multi-Layer Perceptron is introduced into the solution of inverse kinematics process to determine the poses of joints, which achieves precise results than the commonly-used numerical methods in robotics. Besides, a Contact Region Recognition Network (CRRNet) is also proposed to robustly determine the contact regions using a single-view video. Experimental results demonstrate that our method outperforms the state-of-the-art on benchmark BEHAVE. Additionally, our approach shows good portability and can be seamlessly integrated into other methods for optimizations.

Fair Federated Learning under Domain Skew with Local Consistency and Domain Diversity

Federated learning (FL) has emerged as a new paradigm for privacy-preserving collaborative training. Under domain skew, the current FL approaches are biased and face two fairness problems. 1) Parameter Update Conflict: data disparity among clients leads to varying parameter importance and inconsistent update directions. These two disparities cause important parameters to potentially be overwhelmed by unimportant ones of dominant updates. It consequently results in significant performance decreases for lower-performing clients. 2) Model Aggregation Bias: existing FL approaches introduce unfair weight allocation and neglect domain diversity. It leads to biased model convergence objective and distinct performance among domains. We discover a pronounced directional update consistency in Federated Learning and propose a novel framework to tackle above issues. First, leveraging the discovered characteristic, we selectively discard unimportant parameter updates to prevent updates from clients with lower performance overwhelmed by unimportant parameters, resulting in fairer generalization performance. Second, we propose a fair aggregation objective to prevent global model bias towards some domains, ensuring that the global model continuously aligns with an unbiased model. The proposed method is generic and can be combined with other existing FL methods to enhance fairness. Comprehensive experiments on Digits and Office-Caltech demonstrate the high fairness and performance of our method.

Can Far-field Beam Training Be Deployed for Cross-field Beam Alignment in Terahertz UM-MIMO Communications?

Ultra-massive multiple-input multiple-output (UM-MIMO) is the enabler of Terahertz (THz) communications in next-generation wireless networks. In THz UM-MIMO systems, a new paradigm of cross-field communications spanning from near-field to far-field is emerging, since the near-field range expands with higher frequencies and larger array apertures. Precise beam alignment in cross-field is critical but challenging. Specifically, unlike far-field beams that rely only on the angle domain, the incorporation of dual-domain (angle and distance) training significantly increases overhead. A natural question arises of whether far-field beam training can be deployed for cross-field beam alignment. In this paper, this question is answered, by demonstrating that the far-field training enables sufficient signal-to-noise ratio (SNR) in both far- and near-field scenarios, while exciting all channel dimensions. Based on that, we propose a subarray-coordinated hierarchical (SCH) training with greatly reduced overhead. To further obtain high-precision beam designs, we propose a two-phase angle and distance beam estimator (TPBE). Extensive simulations demonstrate the effectiveness of the proposed methods. Compared to near-field exhaustive search, the SCH possesses 0.2\% training overhead. The TPBE achieves 0.01~degrees and 0.02~m estimation root-mean-squared errors for angle and distance. Furthermore, with the estimated beam directions, a near-optimal SNR with 0.11~dB deviation is attained after beam alignment.

LogPrompt: Prompt Engineering Towards Zero-Shot and Interpretable Log Analysis

Automated log analysis is crucial in modern software-intensive systems for ensuring reliability and resilience throughout software maintenance and engineering life cycles. Existing methods perform tasks such as log parsing and log anomaly detection by providing a single prediction value without interpretation. However, given the increasing volume of system events, the limited interpretability of analysis results hinders analysts' trust and their ability to take appropriate actions. Moreover, these methods require substantial in-domain training data, and their performance declines sharply (by up to 62.5%) in online scenarios involving unseen logs from new domains, a common occurrence due to rapid software updates. In this paper, we propose LogPrompt, a novel zero-shot and interpretable log analysis approach. LogPrompt employs large language models (LLMs) to perform zero-shot log analysis tasks via a suite of advanced prompt strategies tailored for log tasks, which enhances LLMs' performance by up to 107.5% compared with simple prompts. Experiments on nine publicly available evaluation datasets across two tasks demonstrate that LogPrompt, despite using no training data, outperforms existing approaches trained on thousands of logs by up to around 50%. We also conduct a human evaluation of LogPrompt's interpretability, with six practitioners possessing over 10 years of experience, who highly rated the generated content in terms of usefulness and readability (averagely 4.42/5). LogPrompt also exhibits remarkable compatibility with open-source and smaller-scale LLMs, making it flexible for practical deployment.

ImDiffusion: Imputed Diffusion Models for Multivariate Time Series Anomaly Detection

Anomaly detection in multivariate time series data is of paramount importance for ensuring the efficient operation of large-scale systems across diverse domains. However, accurately detecting anomalies in such data poses significant challenges. Existing approaches, including forecasting and reconstruction-based methods, struggle to address these challenges effectively. To overcome these limitations, we propose a novel anomaly detection framework named ImDiffusion, which combines time series imputation and diffusion models to achieve accurate and robust anomaly detection. The imputation-based approach employed by ImDiffusion leverages the information from neighboring values in the time series, enabling precise modeling of temporal and inter-correlated dependencies, reducing uncertainty in the data, thereby enhancing the robustness of the anomaly detection process. ImDiffusion further leverages diffusion models as time series imputers to accurately capturing complex dependencies. We leverage the step-by-step denoised outputs generated during the inference process to serve as valuable signals for anomaly prediction, resulting in improved accuracy and robustness of the detection process. We evaluate the performance of ImDiffusion via extensive experiments on benchmark datasets. The results demonstrate that our proposed framework significantly outperforms state-of-the-art approaches in terms of detection accuracy and timeliness. ImDiffusion is further integrated into the real production system in Microsoft and observe a remarkable 11.4% increase in detection F1 score compared to the legacy approach. To the best of our knowledge, ImDiffusion represents a pioneering approach that combines imputation-based techniques with time series anomaly detection, while introducing the novel use of diffusion models to the field.