Federated Learning (FL) enables the collaborative training of a global model while protecting participants’ privacy. Evaluating each participant’s contribution is essential to providing a high-quality model, ensuring fairness, and mitigating potential biases. Most existing contribution evaluation approaches for FL assume that the server has a public validation dataset. However, it is almost impossible to obtain a validation dataset due to privacy concerns. In this article, we propose a Federated Privacy-preserving Representation Synthesis (FedPRS) framework to synthesize a validation dataset for contribution evaluation. The proposed FedPRS framework first transforms each participant’s private validation dataset into its representation. Then, a random-region desensitization strategy is developed to further desensitize the dataset without compromising its utility. The desensitized representation dataset of each participant is collected by the server to evaluate federated contribution, which considers both equity and privacy protection. Moreover, we instantiate and integrate three specific contribution evaluation approaches in this framework. We perform experiments on various FL settings, including independently identically distributed (IID) and non-IID data distributions. Experimental results demonstrate that the contribution evaluation results obtained using the validation dataset synthesized by the FedPRS framework are closely aligned with those obtained using a real, private validation dataset.
Cordycepin (3′-deoxyadenosine, 3′-dA), the flagship nucleoside antibiotic from Cordyceps militaris, exerts potent anti-inflammatory, antimicrobial, and antitumor activity but is rapidly inactivated by human adenosine deaminase (ADA). While prodrugs, ADA inhibitors, and nanocarriers have been pursued to prolong its half-life, the influence of solid form on delivery performance remains unexplored. Here, three polymorphs—anhydrate-I (flake-like), anhydrate-II (rod-like), and a previously unreported monohydrate (fibrillar)—were prepared, characterized (PXRD, TG-DSC, FTIR), and subjected to equilibrium solubility, slurry-conversion, and humidity-sorption mapping. The monohydrate dehydrates at 144 °C and sequentially transforms to anhydrate-I → anhydrate-II (ΔH = −127.5 J g−1), establishing a monotropic relationship between the two anhydrous forms. Solubility displays a bell-shaped profile versus water activity: the monohydrate is stable above aw 0.8, whereas anhydrate-II predominates below aw 0.2. In model immediate-release tablets, anhydrate-II achieves complete dissolution within 10 min, whereas the monohydrate sustains release for 30 min. Hygroscopicity tests show the monohydrate absorbs <6% water up to 75% RH without structural change, whereas anhydrate-I converts to the monohydrate above 63% RH. The quantitative humidity–crystal form–performance correlations provide a rational platform for crystal form selection and the design of stable, efficacious cordycepin solid dosage forms.
