Hongliang Lu

Jingtao He, Hongliang Lu, Xiaoyun Qiu et al.

Vision-Language-Action (VLA) models have demonstrated promising capability in autonomous driving, highlighting the potential of unified multimodal architectures for jointly modeling perception and planning. However, how current VLA-based driving behavior is grounded in visual information remains poorly understood. Existing evaluation protocols mainly focus on aggregate performance metrics, lacking structured and practical diagnostics to quantify visual-behavior dependency. In this work, we introduce a structured multi-level visual perturbation framework to analyze visual-behavior dependency in VLA-based driving models systematically. The framework organizes controlled visual perturbations along three complementary dimensions: channellevel degradation, information-level disruption, and structurelevel modification. We apply it to VLA-based driving systems and evaluate behavioral responses under both open-loop trajectory prediction and interactive closed-loop safety evaluation. Experimental results reveal evaluation-dependent dependency patterns and uneven visual grounding across abstraction levels. These findings call for more structured analyses and principled design of VLA driving models to better understand how visual information shapes behavior and develop safer, more robust systems.

Junjie Yu, Pengrui Lu, Weiye Si et al.

Benchmarks within the OpenClaw ecosystem have thus far evaluated exclusively assistant-level tasks, leaving the academic-level capabilities of OpenClaw largely unexamined. We introduce AcademiClaw, a bilingual benchmark of 80 complex, long-horizon tasks sourced directly from university students' real academic workflows -- homework, research projects, competitions, and personal projects -- that they found current AI agents unable to solve effectively. Curated from 230 student-submitted candidates through rigorous expert review, the final task set spans 25+ professional domains, ranging from olympiad-level mathematics and linguistics problems to GPU-intensive reinforcement learning and full-stack system debugging, with 16 tasks requiring CUDA GPU execution. Each task executes in an isolated Docker sandbox and is scored on task completion by multi-dimensional rubrics combining six complementary techniques, with an independent five-category safety audit providing additional behavioral analysis. Experiments on six frontier models show that even the best achieves only a 55\% pass rate. Further analysis uncovers sharp capability boundaries across task domains, divergent behavioral strategies among models, and a disconnect between token consumption and output quality, providing fine-grained diagnostic signals beyond what aggregate metrics reveal. We hope that AcademiClaw and its open-sourced data and code can serve as a useful resource for the OpenClaw community, driving progress toward agents that are more capable and versatile across the full breadth of real-world academic demands. All data and code are available at https://github.com/GAIR-NLP/AcademiClaw.

Xusen Guo, Xinxi Yang, Mingxing Peng et al.

Developing efficient traffic models is crucial for optimizing modern transportation systems. However, current modeling approaches remain labor-intensive and prone to human errors due to their dependence on manual workflows. These processes typically involve extensive literature reviews, formula tuning, and iterative testing, which often lead to inefficiencies. To address this, we propose TR-Agent, an AI-powered framework that autonomously develops and refines traffic models through a closed-loop, iterative process. We structure the research pipeline into four key stages: idea generation, theory formulation, theory evaluation, and iterative optimization, and implement TR-Agent with four corresponding modules. These modules collaborate to retrieve knowledge from external sources, generate novel hypotheses, implement and debug models, and evaluate their performance on evaluation datasets. Through iteratively feedback and refinement, TR-Agent improves both modeling efficiency and effectiveness. We validate the framework on three representative traffic models: the Intelligent Driver Model (IDM) for car-following behavior, the MOBIL model for lane-changing, and the Lighthill-Whitham-Richards (LWR) speed-density relationship for macroscopic traffic flow modeling. Experimental results show substantial performance gains over the original models. To assess the robustness and generalizability of the improvements, we conduct additional evaluations across multiple real-world datasets, demonstrating consistent performance gains beyond the original development data. Furthermore, TR-Agent produces interpretable explanations for each improvement, enabling researchers to easily verify and extend its results. This makes TR-Agent a valuable assistant for traffic modeling refinement and a promising tool for broader applications in transportation research.

Haotian Wang, Yusong Huang, Zhaonian Kuang et al.

Recent feed-forward models have significantly advanced geometry perception for inferring dense 3D structure from sensor observations. However, its essential capabilities remain fragmented across multiple incompatible paradigms, including online perception, offline reconstruction, multi-modal integration, long-horizon scalability, and metric-scale estimation. We present UniT, a unified model built upon a novel Group Autoregressive Transformer, which reformulates these seemingly disparate capabilities within a single framework. The key idea is to treat groups of sensor observations as the basic autoregressive units and predict the corresponding point maps in an anchor-free and scale-adaptive manner. More specifically, diverse view configurations in both online and offline settings are naturally unified within a single group autoregression process. By varying the group size, online mode operates over multiple autoregressive steps with single-frame groups, whereas offline mode aggregates a multi-frame group in a single forward pass. Meanwhile, a queue-style KV caching mechanism ensures bounded autoregressive memory over long horizons. This is enabled by reducing long-range dependencies on early frames through anchor-free relational modeling, thereby allowing outdated memory to be discarded on the fly. To improve metric-scale generalization across scenes, a scale-adaptive geometry loss is further introduced within this framework. It couples relative geometric constraints with a partial absolute scale term, implicitly regularizing global scale and inducing a progressive transition from scale-invariant geometry to metric-scale solutions. Together with a dedicated modal attention module for integrating auxiliary modalities, UniT achieves state-of-the-art performance in unified geometry perception, as validated on ten benchmarks spanning seven representative tasks.

Hongliang Lu, Zhonglin Xie, Yaoyu Wu et al. · pku

Despite the rapid development of large language models (LLMs), a fundamental challenge persists: the lack of high-quality optimization modeling datasets hampers LLMs' robust modeling of practical optimization problems from natural language descriptions (NL). This data scarcity also contributes to the generalization difficulties experienced by learning-based methods. To address these challenges, we propose a scalable framework for synthesizing a high-quality dataset, named OptMATH. Starting from curated seed data with mathematical formulations (MF), this framework automatically generates problem data (PD) with controllable complexity. Then, a back-translation step is employed to obtain NL. To verify the correspondence between the NL and the PD, a forward modeling step followed by rejection sampling is used. The accepted pairs constitute the training part of OptMATH. Then a collection of rejected pairs is identified and further filtered. This collection serves as a new benchmark for optimization modeling, containing difficult instances whose lengths are much longer than these of NL4OPT and MAMO. Through extensive experiments, we demonstrate that models of various sizes (0.5B-32B parameters) trained on OptMATH achieve superior results on multiple modeling benchmarks, thereby validating the effectiveness and scalability of our approach. Our dataset is publicly available at https://github.com/AuroraLHL/OptMATH.

Tianshun Li, Hongliang Lu, Yanggang Sheng et al.

Ensuring energy feasibility under wind uncertainty is critical for the safety and reliability of UAV delivery missions. In realistic truck-drone logistics systems, UAVs must deliver parcels and safely return under time-varying wind conditions that are only partially observable during flight. However, most existing routing approaches assume static or deterministic energy models, making them unreliable in dynamic wind environments. We propose Battery-Efficient Routing (BER), an online risk-sensitive planning framework for wind-sensitive truck-assisted UAV delivery. The problem is formulated as routing on a time dependent energy graph whose edge costs evolve according to wind-induced aerodynamic effects. BER continuously evaluates return feasibility while balancing instantaneous energy expenditure and uncertainty-aware risk. The approach is embedded in a hierarchical aerial-ground delivery architecture that combines task allocation, routing, and decentralized trajectory execution. Extensive simulations on synthetic ER graphs generated in Unreal Engine environments and quasi-real wind logs demonstrate that BER significantly improves mission success rates and reduces wind-induced failures compared with static and greedy baselines. These results highlight the importance of integrating real-time energy budgeting and environmental awareness for UAV delivery planning under dynamic wind conditions.

Ziyu Wang, Chenyuan Liu, Yushun Xiang et al.

Recently, with the rapid development of robot learning and imitation learning, numerous datasets and methods have emerged. However, these datasets and their task designs often lack systematic consideration and principles. This raises important questions: Do the current datasets and task designs truly advance the capabilities of robotic agents? Do evaluations on a few common tasks accurately reflect the differentiated performance of various methods proposed by different teams and evaluated on different tasks? To address these issues, we introduce the Great March 100 (\textbf{GM-100}) as the first step towards a robot learning Olympics. GM-100 consists of 100 carefully designed tasks that cover a wide range of interactions and long-tail behaviors, aiming to provide a diverse and challenging set of tasks to comprehensively evaluate the capabilities of robotic agents and promote diversity and complexity in robot dataset task designs. These tasks are developed through systematic analysis and expansion of existing task designs, combined with insights from human-object interaction primitives and object affordances. We collect a large amount of trajectory data on different robotic platforms and evaluate several baseline models. Experimental results demonstrate that the GM-100 tasks are 1) feasible to execute and 2) sufficiently challenging to effectively differentiate the performance of current VLA models. Our data and code are available at https://rhos.ai/research/gm-100.

Hui Zhong, Xianda Chen, PakHin Tiu et al.

To alleviate energy shortages and environmental impacts caused by transportation, this study introduces EcoFollower, a novel eco-car-following model developed using reinforcement learning (RL) to optimize fuel consumption in car-following scenarios. Employing the NGSIM datasets, the performance of EcoFollower was assessed in comparison with the well-established Intelligent Driver Model (IDM). The findings demonstrate that EcoFollower excels in simulating realistic driving behaviors, maintaining smooth vehicle operations, and closely matching the ground truth metrics of time-to-collision (TTC), headway, and comfort. Notably, the model achieved a significant reduction in fuel consumption, lowering it by 10.42\% compared to actual driving scenarios. These results underscore the capability of RL-based models like EcoFollower to enhance autonomous vehicle algorithms, promoting safer and more energy-efficient driving strategies.

Xiao Liu, Kai Liu, Naiyang Guan et al.

Diffusion and rectified flow (RF) models generate high-fidelity images and videos, but their iterative velocity-field evaluations are computationally expensive. Existing caching methods accelerate sampling by skipping timesteps, yet their coarse approximations introduce accumulated errors over long skip intervals and degrade quality under aggressive acceleration. We propose TACache (Trajectory-Aware Cache), a training-free acceleration framework following a skip-then-compensate paradigm. TACache performs an orthogonal decomposition of discrete velocity acceleration along the RF trajectory into a parallel component and an orthogonal residual, isolating the magnitude and directional sources of per-step approximation error. The framework operates in two stages: offline, cumulative variation thresholds on the magnitude and direction indicators yield the skip schedule and bound how far each skip interval may extend; online, at each skipped step the offline statistics are combined with the sample's historical orthogonal direction to reconstruct the skipped velocity without additional model evaluations. Experiments on BAGEL, FLUX.1-dev, and Wan2.1-1.3B show that TACache achieves up to 4.14 speedup on text-to-image generation and 2.11 speedup on text-to-video generation, with consistent improvements over prior cache-based methods on all reference-based fidelity metrics. Code will be released soon.

Hongliang Lu, Yuhang Wen, Pengyu Cheng et al. · pku

Reinforcement learning with verifiable rewards (RLVR) has become the mainstream technique for training LLM agents. However, RLVR highly depends on well-crafted task queries and corresponding ground-truth answers to provide accurate rewards, which requires massive human efforts and hinders the RL scaling processes, especially under agentic scenarios. Although a few recent works explore task synthesis methods, the difficulty of generated agentic tasks can hardly be controlled to provide effective RL training advantages. To achieve agentic RLVR with higher scalability, we explore self-play training for deep search agents, in which the learning LLM utilizes multi-turn search engine calling and acts simultaneously as both a task proposer and a problem solver. The task proposer aims to generate deep search queries with well-defined ground-truth answers and increasing task difficulty. The problem solver tries to handle the generated search queries and output the correct answer predictions. To ensure that each generated search query has accurate ground truth, we collect all the searching results from the proposer's trajectory as external knowledge, then conduct retrieval-augmentation generation (RAG) to test whether the proposed query can be correctly answered with all necessary search documents provided. In this search self-play (SSP) game, the proposer and the solver co-evolve their agent capabilities through both competition and cooperation. With substantial experimental results, we find that SSP can significantly improve search agents' performance uniformly on various benchmarks without any supervision under both from-scratch and continuous RL training setups. The code is at https://github.com/Alibaba-Quark/SSP.

Xusen Guo, Mingxing Peng, Hongliang Lu et al.

Participatory urban sensing leverages human mobility for large-scale urban data collection, yet existing methods typically rely on centralized optimization and assume homogeneous participants, resulting in rigid assignments that overlook personal preferences and heterogeneous urban contexts. We propose MAPUS, an LLM-based multi-agent framework for personalized and fair participatory urban sensing. In our framework, participants are modeled as autonomous agents with individual profiles and schedules, while a coordinator agent performs fairness-aware selection and refines sensing routes through language-based negotiation. Experiments on real-world datasets show that MAPUS achieves competitive sensing coverage while substantially improving participant satisfaction and fairness, promoting more human-centric and sustainable urban sensing systems.

Ziyi Shi, Xusen Guo, Hongliang Lu et al.

Effective pandemic control requires timely and coordinated policymaking across administrative regions that are intrinsically interdependent. However, human-driven responses are often fragmented and reactive, with policies formulated in isolation and adjusted only after outbreaks escalate, undermining proactive intervention and global pandemic mitigation. To address this challenge, here we propose a large language model (LLM) multi-agent policymaking framework that supports coordinated and proactive pandemic control across regions. Within our framework, each administrative region is assigned an LLM agent as an AI policymaking assistant. The agent reasons over region-specific epidemiological dynamics while communicating with other agents to account for cross-regional interdependencies. By integrating real-world data, a pandemic evolution simulator, and structured inter-agent communication, our framework enables agents to jointly explore counterfactual intervention scenarios and synthesize coordinated policy decisions through a closed-loop simulation process. We validate the proposed framework using state-level COVID-19 data from the United States between April and December 2020, together with real-world mobility records and observed policy interventions. Compared with real-world pandemic outcomes, our approach reduces cumulative infections and deaths by up to 63.7% and 40.1%, respectively, at the individual state level, and by 39.0% and 27.0%, respectively, when aggregated across states. These results demonstrate that LLM multi-agent systems can enable more effective pandemic control with coordinated policymaking...

Xun Zhang, Kaicheng Yang, Hongliang Lu et al.

Recently, Diffusion Transformers (DiTs) have emerged in Real-World Image Super-Resolution (Real-ISR) to generate high-quality textures, yet their heavy inference burden hinders real-world deployment. While Post-Training Quantization (PTQ) is a promising solution for acceleration, existing methods in super-resolution mostly focus on U-Net architectures, whereas generic DiT quantization is typically designed for text-to-image tasks. Directly applying these methods to DiT-based super-resolution models leads to severe degradation of local textures. Therefore, we propose Q-DiT4SR, the first PTQ framework specifically tailored for DiT-based Real-ISR. We propose H-SVD, a hierarchical SVD that integrates a global low-rank branch with a local block-wise rank-1 branch under a matched parameter budget. We further propose Variance-aware Spatio-Temporal Mixed Precision: VaSMP allocates cross-layer weight bit-widths in a data-free manner based on rate-distortion theory, while VaTMP schedules intra-layer activation precision across diffusion timesteps via dynamic programming (DP) with minimal calibration. Experiments on multiple real-world datasets demonstrate that our Q-DiT4SR achieves SOTA performance under both W4A6 and W4A4 settings. Notably, the W4A4 quantization configuration reduces model size by 5.8$\times$ and computational operations by over 60$\times$. Our code and models will be available at https://github.com/xunzhang1128/Q-DiT4SR.

Xianda Chen, Meixin Zhu, Kehua Chen et al.

Car-following is a control process in which a following vehicle (FV) adjusts its acceleration to keep a safe distance from the lead vehicle (LV). Recently, there has been a booming of data-driven models that enable more accurate modeling of car-following through real-world driving datasets. Although there are several public datasets available, their formats are not always consistent, making it challenging to determine the state-of-the-art models and how well a new model performs compared to existing ones. In contrast, research fields such as image recognition and object detection have benchmark datasets like ImageNet, Microsoft COCO, and KITTI. To address this gap and promote the development of microscopic traffic flow modeling, we establish a public benchmark dataset for car-following behavior modeling. The benchmark consists of more than 80K car-following events extracted from five public driving datasets using the same criteria. These events cover diverse situations including different road types, various weather conditions, and mixed traffic flows with autonomous vehicles. Moreover, to give an overview of current progress in car-following modeling, we implemented and tested representative baseline models with the benchmark. Results show that the deep deterministic policy gradient (DDPG) based model performs competitively with a lower MSE for spacing compared to traditional intelligent driver model (IDM) and Gazis-Herman-Rothery (GHR) models, and a smaller collision rate compared to fully connected neural network (NN) and long short-term memory (LSTM) models in most datasets. The established benchmark will provide researchers with consistent data formats and metrics for cross-comparing different car-following models, promoting the development of more accurate models. We open-source our dataset and implementation code in https://github.com/HKUST-DRIVE-AI-LAB/FollowNet.

Zhong Li, Hongliang Lu, Tao Wei et al.

Optimization modeling underpins decision-making in logistics, manufacturing, energy, and finance, yet translating natural-language requirements into correct optimization formulations and solver-executable code remains labor-intensive. Although large language models (LLMs) have been explored for this task, evaluation is still dominated by toy-sized or synthetic benchmarks, masking the difficulty of industrial problems with $10^{3}$--$10^{6}$ (or more) variables and constraints. A key bottleneck is the lack of benchmarks that align natural-language specifications with reference formulations/solver code grounded in real optimization models. To fill in this gap, we introduce MIPLIB-NL, built via a structure-aware reverse construction methodology from real mixed-integer linear programs in MIPLIB~2017. Our pipeline (i) recovers compact, reusable model structure from flat solver formulations, (ii) reverse-generates natural-language specifications explicitly tied to this recovered structure under a unified model--data separation format, and (iii) performs iterative semantic validation through expert review and human--LLM interaction with independent reconstruction checks. This yields 223 one-to-one reconstructions that preserve the mathematical content of the original instances while enabling realistic natural-language-to-optimization evaluation. Experiments show substantial performance degradation on MIPLIB-NL for systems that perform strongly on existing benchmarks, exposing failure modes invisible at toy scale.

Hongliang Lu, Yunmeng Liu, Junjie Yang

Human decision-making heavily relies on active sensing, a well-documented cognitive behaviour for evidence gathering to accommodate ever-changing environments. However, its operational mechanism in the real world remains non-trivial. Currently, an in-laboratory paradigm, called evidence accumulation modelling (EAM), points out that human decision-making involves transforming external evidence into internal mental beliefs. However, the gap in evidence affordance between real-world contexts and laboratory settings hinders the effective application of EAM. Here we generalize EAM to the real world and conduct analysis in real-world driving scenarios. A cognitive scheme is proposed to formalize real-world evidence affordance and capture active sensing through eye movements. Empirically, our scheme can plausibly portray the accumulation of drivers' mental beliefs, explaining how active sensing transforms evidence into mental beliefs from the perspective of information utility. Also, our results demonstrate a negative correlation between evidence affordance and attention recruited by individuals, revealing how human drivers adapt their evidence-collection patterns across various contexts. Moreover, we reveal the positive influence of evidence affordance and attention distribution on decision-making propensity. In a nutshell, our computational scheme generalizes EAM to real-world contexts and provides a comprehensive account of how active sensing underlies real-world decision-making, unveiling multifactorial, integrated characteristics in real-world decision-making.

Pei Liu, Hongliang Lu, Haichao Liu et al.

Human vision is capable of transforming two-dimensional observations into an egocentric three-dimensional scene understanding, which underpins the ability to translate complex scenes and exhibit adaptive behaviors. This capability, however, remains lacking in current autonomous driving systems, where mainstream approaches primarily rely on depth-based 3D reconstruction rather than true scene understanding. To address this limitation, we propose a novel human-like framework called OmniScene. First, we introduce the OmniScene Vision-Language Model (OmniVLM), a vision-language framework that integrates multi-view and temporal perception for holistic 4D scene understanding. Then, harnessing a teacher-student OmniVLM architecture and knowledge distillation, we embed textual representations into 3D instance features for semantic supervision, enriching feature learning, and explicitly capturing human-like attentional semantics. These feature representations are further aligned with human driving behaviors, forming a more human-like perception-understanding-action architecture. In addition, we propose a Hierarchical Fusion Strategy (HFS) to address imbalances in modality contributions during multimodal integration. Our approach adaptively calibrates the relative significance of geometric and semantic features at multiple abstraction levels, enabling the synergistic use of complementary cues from visual and textual modalities. This learnable dynamic fusion enables a more nuanced and effective exploitation of heterogeneous information. We evaluate OmniScene comprehensively on the nuScenes dataset, benchmarking it against over ten state-of-the-art models across various tasks. Our approach consistently achieves superior results, establishing new benchmarks in perception, prediction, planning, and visual question answering.

Hongliang Lu, Shuqi Shen, Junjie Yang et al.

More than the adherence to specific traffic regulations, driving culture touches upon a more implicit part - an informal, conventional, collective behavioral pattern followed by drivers - that varies across countries, regions, and even cities. Such cultural divergence has become one of the biggest challenges in deploying autonomous vehicles (AVs) across diverse regions today. The current emergence of data-driven methods has shown a potential solution to enable culture-compatible driving through learning from data, but what if some underdeveloped regions cannot provide sufficient local data to inform driving culture? This issue is particularly significant for a broader global AV market. Here, we propose a cross-cultural deployment scheme for AVs, called data-light inverse reinforcement learning, designed to re-calibrate culture-specific AVs and assimilate them into other cultures. First, we report the divergence in driving cultures through a comprehensive comparative analysis of naturalistic driving datasets on highways from three countries: Germany, China, and the USA. Then, we demonstrate the effectiveness of our scheme by testing the expeditious cross-cultural deployment across these three countries, with cumulative testing mileage of over 56084 km. The performance is particularly advantageous when cross-cultural deployment is carried out without affluent local data. Results show that we can reduce the dependence on local data by a margin of 98.67% at best. This study is expected to bring a broader, fairer AV global market, particularly in those regions that lack enough local data to develop culture-compatible AVs.