Multimodal Large Language Models are primarily trained and evaluated on aligned image-text pairs, which leaves their ability to detect and resolve real-world inconsistencies largely unexplored. In open-domain applications visual and textual cues often conflict, requiring models to perform structured reasoning beyond surface-level alignment. We introduce CrossCheck-Bench, a diagnostic benchmark for evaluating contradiction detection in multimodal inputs. The benchmark adopts a hierarchical task framework covering three levels of reasoning complexity and defines seven atomic capabilities essential for resolving cross-modal inconsistencies. CrossCheck-Bench includes 15k question-answer pairs sourced from real-world artifacts with synthetically injected contradictions. The dataset is constructed through a multi-stage annotation pipeline involving more than 450 expert hours to ensure semantic validity and calibrated difficulty across perception, integration, and reasoning. We evaluate 13 state-of-the-art vision-language models and observe a consistent performance drop as tasks shift from perceptual matching to logical contradiction detection. Most models perform well on isolated entity recognition but fail when multiple clues must be synthesized for conflict reasoning. Capability-level analysis further reveals uneven skill acquisition, especially in tasks requiring multi-step inference or rule-based validation. Additional probing shows that conventional prompting strategies such as Chain-of-Thought and Set-of-Mark yield only marginal gains. By contrast, methods that interleave symbolic reasoning with grounded visual processing achieve more stable improvements. These results highlight a persistent bottleneck in multimodal reasoning and suggest new directions for building models capable of robust cross-modal verification.

Safe Reinforcement Learning (RL) often faces significant issues such as constraint violations and instability, necessitating the use of constrained policy optimization, which seeks optimal policies while ensuring adherence to specific constraints like safety. Typically, constrained optimization problems are addressed by the Lagrangian method, a post-violation remedial approach that may result in oscillations and overshoots. Motivated by this, we propose a novel method named Proactive Constrained Policy Optimization (PCPO) that incorporates a preemptive penalty mechanism. This mechanism integrates barrier items into the objective function as the policy nears the boundary, imposing a cost. Meanwhile, we introduce a constraint-aware intrinsic reward to guide boundary-aware exploration, which is activated only when the policy approaches the constraint boundary. We establish theoretical upper and lower bounds for the duality gap and the performance of the PCPO update, shedding light on the method's convergence characteristics. Additionally, to enhance the optimization performance, we adopt a policy iteration approach. An interesting finding is that PCPO demonstrates significant stability in experiments. Experimental results indicate that the PCPO framework provides a robust solution for policy optimization under constraints, with important implications for future research and practical applications.

Dongxu Zhang, Yiding Sun, Cheng Tan, Wenbiao Yan, Ning Yang, Jihua Zhu, Haijun Zhang. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Mobile Edge Computing (MEC) leverages computational heterogeneity between mobile devices and edge nodes to enable real-time applications requiring high information fresh ness. The Age-of-Information (AoI) metric serves as a crucial evaluator of information timeliness in such systems. Addressing AoI minimization in multi-user MEC environments presents significant challenges due to stochastic computing times. In this paper, we consider multiple users offloading tasks to heterogeneous edge servers in an MEC system, focusing on preemptive and non-preemptive task scheduling mechanisms. The problem is first reformulated as a Restless Multi-Arm Bandit (RMAB) problem, with a multi-layer Markov Decision Process (MDP) framework established to characterize AoI dynamics in the MEC system. Based on the multi-layer MDP, we propose a nested index framework and design a nested index policy with provably asymptotic optimality. This establishes a theoretical framework adaptable to various scheduling mechanisms, achieving efficient optimization through state stratification and index design in both preemptive and non-preemptive modes. Finally, the closed-form of the nested index is derived, facilitating performance trade-offs between computational complexity and accuracy while ensuring the universal applicability of the nested index policy across both scheduling modes. The experimental results show that in non preemptive scheduling, compared with the benchmark method, the optimality gap is reduced by 25.43%, while in preemptive scheduling, the gap has reduced by 21.21%. As the system scale increases, it asymptotically converges in two scheduling modes and especially provides near-optimal performance in a non preemptive structure.

This paper presents a Long Short-Term Memory network-based Fluid Experiment Data-Driven model (FEDLSTM) for predicting unsteady, nonlinear hydrodynamic forces on the underwater quadruped robot we constructed. Trained on experimental data from leg force and body drag tests conducted in both a recirculating water tank and a towing tank, FED-LSTM outperforms traditional Empirical Formulas (EF) commonly used for flow prediction over flat surfaces. The model demonstrates superior accuracy and adaptability in capturing complex fluid dynamics, particularly in straightline and turning-gait optimizations via the NSGA-II algorithm. FED-LSTM reduces deflection errors during straight-line swimming and improves turn times without increasing the turning radius. Hardware experiments further validate the model's precision and stability over EF. This approach provides a robust framework for enhancing the swimming performance of legged robots, laying the groundwork for future advances in underwater robotic locomotion.

Mobile edge computing (MEC) is essential for next-generation mobile network applications that prioritize various performance metrics, including delays and energy efficiency. However, conventional single-objective scheduling solutions cannot be directly applied to practical systems in which the preferences (i.e., the weights of different objectives) are often unknown or challenging to specify in advance. In this study, we formulate a multi-objective offloading problem for MEC with multiple edges to minimize the sum of expected long-term energy consumption and delay while considering unknown preferences. To address the challenge of unknown preferences and the potentially diverse MEC systems, we propose a generalizable multi-objective (deep) reinforcement learning (GMORL)-based tasks offloading framework, which employs the Discrete Soft Actor-Critic (Discrete-SAC) method. Our method uses a single policy model to efficiently schedule tasks based on varying preferences and adapt to heterogeneous MEC systems with different CPU frequencies and server quantities. Under the proposed framework, we introduce a histogram-based state encoding method for constructing features for multiple edges in MEC systems, a sophisticated reward function for accurately computing the utilities of delay and energy consumption, and a novel neural network architecture for improving generalization. Simulation results demonstrate that our proposed GMORL scheme enhances the hypervolume of the Pareto front by up to 121.0% compared to benchmarks.

Fine-tuning pre-trained Large Language Models (LLMs) is essential to align them with human values and intentions. This process often utilizes methods like pairwise comparisons and KL divergence against a reference LLM, focusing on the evaluation of full answers generated by the models. However, the generation of these responses occurs in a token level, following a sequential, auto-regressive fashion. In this paper, we introduce Token-level Direct Preference Optimization (TDPO), a novel approach to align LLMs with human preferences by optimizing policy at the token level. Unlike previous methods, which face challenges in divergence efficiency, TDPO integrates forward KL divergence constraints for each token, improving alignment and diversity. Utilizing the Bradley-Terry model for a token-based reward system, our method enhances the regulation of KL divergence, while preserving simplicity without the need for explicit reward modeling. Experimental results across various text tasks demonstrate TDPO’s superior performance in balancing alignment with generation diversity. Notably, fine-tuning with TDPO strikes a better balance than DPO in the controlled sentiment generation and single-turn dialogue datasets, and significantly improves the quality of generated responses compared to both DPO and PPO-based RLHF methods.

Exploiting the computational heterogeneity of mobile devices and edge nodes, mobile edge computation (MEC) provides an efficient approach to achieving real-time applications that are sensitive to information freshness, by offloading tasks from mobile devices to edge nodes. We use the metric Age-of-Information (AoI) to evaluate information freshness. An efficient solution to minimize the AoI for the MEC system with multiple users is non-trivial to obtain due to the random computing time. In this paper, we consider multiple users offloading tasks to heterogeneous edge servers in a MEC system. We first reformulate the problem as a Restless Multi-Arm-Bandit (RMAB) problem and establish a hierarchical Markov Decision Process (MDP) to characterize the updating of AoI for the MEC system. Based on the hierarchical MDP, we propose a nested index framework and design a nested index policy with provably asymptotic optimality. Finally, the closed form of the nested index is obtained, which enables the performance tradeoffs between computation complexity and accuracy. Our algorithm leads to an optimality gap reduction of up to 40%, compared to benchmarks. Our algorithm asymptotically approximates the lower bound as the system scalar gets large enough.

Mobile edge computing (MEC) is essential for next-generation mobile network applications that prioritize various performance metrics, including delays and energy consumption. However, conventional single-objective scheduling solutions cannot be directly applied to practical systems in which the preferences of these applications (i.e., the weights of different objectives) are often unknown or challenging to specify in advance. In this study, we address this issue by formulating a multi-objective offloading problem for MEC with multiple edges to minimize expected long-term energy consumption and transmission delay while considering unknown preferences as parameters. To address the challenge of unknown preferences, we design a multi-objective (deep) reinforcement learning (MORL)-based resource scheduling scheme with proximal policy optimization (PPO). In addition, we introduce a well-designed state encoding method for constructing features for multiple edges in MEC systems, a sophisticated reward function for accurately computing the utilities of delay and energy consumption. Simulation results demonstrate that our proposed MORL scheme enhances the hypervolume of the Pareto front by up to 233.1% compared to benchmarks.

Joint caching and transmission optimization problem is challenging due to the deep coupling between decisions. This paper proposes an iterative distributed multi-agent learning approach to jointly optimize caching and transmission. The goal of this approach is to minimize the total transmission delay of all users. In this iterative approach, each iteration includes caching optimization and transmission optimization. A multi-agent reinforcement learning (MARL)-based caching network is developed to cache popular tasks, such as answering which files to evict from the cache and which files to storage. Based on the cached files of the caching network, the transmission network transmits cached files for users by single transmission (ST) or joint transmission (JT) with multi-agent Bayesian learning automaton (MABLA) method. And then users access the edge servers with the minimum transmission delay. The experimental results demonstrate the performance of the proposed multi-agent learning approach.

In this paper, the problem of dynamic resource management in a cognitive radio network (CRN) with multiple primary users (PUs), multiple secondary users (SUs), and multiple channels is investigated. An optimization problem is formulated as a multi-agent partially observable Markov decision process (POMDP) problem in a dynamic and not fully observable environment. We consider using deep reinforcement learning (DRL) to address this problem. Based on the channel occupancy of PUs, a multi-agent deep Q-network (DQN)-based dynamic joint spectrum access and mode selection (SAMS) scheme is proposed for the SUs in the partially observable environment. The current observation of each SU is mapped to a suitable action. Each secondary user (SU) takes its own decision without exchanging information with other SUs. It seeks to maximize the total sum rate. Simulation results verify the effectiveness of our proposed schemes.

In the current researches, artificial intelligence (AI) plays a crucial role in resource management for the next generation wireless communication network. However, traditional RL cannot solve the continuous and high dimensional problems. To handle these problems, the concept of deep neural network (DNN) is introduced into RL to solve high dimensional problems. In this paper, we first construct an information interaction model among primary user (PU), secondary user (SU) and wireless sensors in a cognitive radio system. In the model, the SU is unable to get the power allocation information of the PU, and needs to use the received signal strengths (RSSs) of the wireless sensors to adjust its own power. The PU allocates transmit power relying on its power control scheme. We propose an asynchronous advantage actor critic (A3C)-based power control of SU that is a parallel actor-learners framework with root mean square prop (RMSProp) optimization. Multiple SUs learn power control scheme simultaneously on different CPU threads, reducing neural network gradient update interdependence. To further improve the efficiency of spectrum sharing, the distributed proximal policy optimization (DPPO)-based power control is proposed which is an asynchronous variant of actor-critic with adaptive moment (Adam) optimization. It enables the network to converge quickly. After several power adjustments, the PU and the SU meet quality of service (QoS) requirements and achieve spectrum sharing.

In this paper, we investigate the joint subcarrier (SC) assignment and power allocation problem for non-orthogonal multiple access (NOMA) amplify-and- forward two-way relay wireless networks. We aim to maximize the achievable secrecy energy efficiency by jointly designing the SC assignment, user pair scheduling and power allocation. Assuming the perfect knowledge of the channel state information (CSI) at the relay station, we propose a low-complexity subcarrier assignment scheme (SCAS-1), which is equivalent to many-to-many matching games, and then SCAS-2 is formulated as a secrecy energy efficiency maximization problem. The secure power allocation problem is modeled as a convex geometric programming (GP) problem, and then solved by interior point methods. Simulation results demonstrate that the effectiveness of the proposed SSPA algorithms.

Secure communication is a promising technology for wireless networks because it ensures secure transmission of information. In this paper, we investigate the joint subcarrier (SC) assignment and power allocation problem for non-orthogonal multiple access amplify-and-forward two-way relay wireless networks, in the presence of eavesdroppers. By exploiting cooperative jamming (CJ) to enhance the security of the communication link, we aim to maximize the achievable secrecy energy efficiency by jointly designing the SC assignment, user pair scheduling and power allocation. Assuming the perfect knowledge of the channel state information at the relay station, we propose a low-complexity subcarrier assignment scheme (SCAS-1), which is equivalent to many-to-many matching games, and then SCAS-2 is formulated as a secrecy energy efficiency maximization problem. The secure power allocation problem is modeled as a convex geometric programming problem, and then, solved by interior point methods. Simulation results demonstrate that the effectiveness of the proposed SSPA algorithms under scenarios of using and not using CJ, respectively.