Qixun Zhang

Yirong Zeng, Yufei Liu, Xiao Ding et al.

A central belief in scaling reinforcement learning with verifiable rewards for instruction following (IF) tasks is that, a diverse mixture of verifiable hard and unverifiable soft constraints is essential for generalizing to unseen instructions. In this work, we challenge this prevailing consensus through a systematic empirical investigation. Counter-intuitively, we find that models trained on hard-only constraints consistently outperform those trained on mixed datasets. Extensive experiments reveal that reward precision, rather than constraint diversity, is the primary driver of effective alignment. The LLM judge suffers from a low recall rate in detecting false response, which leads to severe reward hacking, thereby undermining the benefits of diversity. Furthermore, analysis of the attention mechanism reveals that high-precision rewards develop a transferable meta-skill for IF. Motivated by these insights, we propose a simple yet effective data-centric refinement strategy that prioritizes reward precision. Evaluated on five benchmarks, our approach outperforms competitive baselines by 13.4\% in performance while achieving a 58\% reduction in training time, maintaining strong generalization beyond instruction following. Our findings advocate for a paradigm shift: moving away from the indiscriminate pursuit of data diversity toward high-precision rewards.

Kan Yu, Kaixuan Li, Yujia Zhao et al.

The rapid proliferation of unmanned aerial vehicle (UAV) applications imposes stringent requirements on continuous and reliable communication coverage in low-altitude airspace. Conventional cellular systems built upon fixed-position antennas (FPAs) are inherently constrained by static array geometries and limited mechanical degrees of freedom, which severely restrict their ability to adapt to highly dynamic three-dimensional (3D) propagation environments. Movable antenna (MA) technology has recently emerged as a promising paradigm to overcome these limitations by actively reconfiguring electromagnetic radiation characteristics through controllable antenna positioning and array orientation, thereby enabling flexible spatial coverage adaptation. To systematically quantify the airspace coverage capability of MA-enabled systems, this paper formulates a spatial coverage maximization problem over a discretized 3D voxel space. For each voxel, the received signal-to-noise ratio (SNR) is maximized via joint optimization of the MA's 3D positions and beamforming matrices. To efficiently solve the resulting non-convex problem, a hybrid particle swarm optimization and simulated annealing framework is developed to search for high-quality antenna configurations. Simulation results demonstrate that the proposed MA design framework substantially outperforms conventional FPA-based schemes in terms of spatial coverage, achieving coverage rates of 26.8% and 29.65% for airspace below 300m and 600m, respectively. Moreover, further coverage enhancement can be attained by incorporating mechanical tilt adjustment, highlighting the strong potential of MA technology for reliable low-altitude communication coverage.

Yang Luo, Qixun Zhang, Qingni Shen et al.

With the expansion of the market share occupied by the Android platform, security issues (especially application security) have become attention focus of researchers. In fact, the existing methods lack the capabilities to manage application permissions without root privilege. This study proposes a dynamic management mechanism of Android application permissions based on security policies. The paper first describes the permissions by security policies, then implementes permission checking code and request evaluation algorithm in Android framework layer. Experimental results indicate that the presented approach succeeds in permission management of Android applications, and its system overhead is low, which makes it an effective method for Android permission management.