Constant stress creep experiments at 550 °C were performed on a high-purity Fe-3Au-4W (wt.%) ternary alloy with about 1 at.% supersaturation for Au and W in order to study self healing of grain-boundary cavities by both Au-rich and W-rich precipitates. Using synchrotron X-ray nano-tomography, the development of the creep cavities and the healing precipitates at different stages of creep was visualised using two spatial resolutions (30 and 100 nm voxel size) for separate samples taken after different loading times. The healing kinetics was found to strongly depend on the nucleation time of the cavities. Cavities nucleated at an early stage of creep could be fully healed, while the healing of the late-nucleated cavities is much slower due to a decrease in the diffusional flux of the healing supersaturated solutes over time, as a result of (i) a decrease in inter-cavity spacing caused by cavity nucleation and (ii) a gradual depletion of the supersaturated solutes near the grain boundaries. The interaction between the competing healing mechanisms for creep cavities by Au-rich and W-rich precipitates is discussed. It was found that Au-rich precipitates are formed much faster than the W-rich precipitates, and thereby effectively provide creep damage healing on different time scales.

With tremendous amount of recommendation algorithms proposed every year, one critical issue has attracted a considerable amount of attention: there are no effective benchmarks for evaluation, which leads to two major concerns, i.e., unreproducible evaluation and unfair comparison. This paper aims to conduct rigorous (i.e., reproducible and fair) evaluation for implicit-feedback based top-N recommendation algorithms. We first systematically review 85 recommendation papers published at eight top-tier conferences (e.g., RecSys, SIGIR) to summarize important evaluation factors, e.g., data splitting and parameter tuning strategies, etc. Through a holistic empirical study, the impacts of different factors on recommendation performance are then analyzed in-depth. Following that, we create benchmarks with standardized procedures and provide the performance of seven well-tuned state-of-the-arts across six metrics on six widely-used datasets as a reference for later study. Additionally, we release a user-friendly Python toolkit, which differs from existing ones in addressing the broad scope of rigorous evaluation for recommendation. Overall, our work sheds light on the issues in recommendation evaluation and lays the foundation for further investigation. Our code and datasets are available at GitHub (https://github.com/AmazingDD/daisyRec).