Multiple Instance Learning (MIL) is a standard weakly supervised approach for Whole Slide Image (WSI) classification, where performance hinges on both feature representation and MIL pooling strategies. Recent research has predominantly focused on Transformer-based architectures adapted for WSIs. However, we argue that this trend faces a fundamental limitation: data scarcity. In typical settings, Transformer models yield only marginal gains without access to large-scale datasets—resources that are virtually inaccessible to all but a few well-funded research labs. Motivated by this, we revisit simple, non-attention MIL with unsupervised slide features and analyze temperature-$\beta$-controlled log-sum-exp (LSE) pooling. For slides partitioned into $N$ patches, we theoretically show that LSE has a smooth transition at a critical $\beta_{\mathrm{crit}}=\mathcal{O}(\log N)$ threshold, interpolating between mean-like aggregation (stable, better generalization but less sensitive) and max-like aggregation (more sensitive but looser generalization bounds). Grounded in this analysis, we introduce Maxsoft—a novel MIL pooling function that enables flexible control over this trade-off, allowing adaptation to specific tasks and datasets. To further tackle real-world deployment challenges such as specimen heterogeneity, we propose PerPatch augmentation—a simple yet effective technique that enhances model robustness. Empirically, Maxsoft achieves state-of-the-art performance in low-data regimes across four major benchmarks (CAMELYON16, CAMELYON17, TCGA-Lung, and SICAP-MIL), often matching or surpassing large-scale foundation models. When combined with PerPatch augmentation, this performance is further improved through increased robustness. Code is available at \href{https://github.com/jafarinia/maxsoft}{\texttt{https://github.com/jafarinia/maxsoft}}

Self-stabilization in distributed systems is a technique to guarantee convergence to a set of legitimate states without external intervention when a transient fault or bad initialization occurs. Recently, there has been a surge of efforts in designing techniques for automated synthesis of self-stabilizing algorithms that are correct by construction. Most of these techniques, however, are not parameterized, meaning that they can only synthesize a solution for a fixed and predetermined number of processes. In this paper, we report a breakthrough in parameterized synthesis of self-stabilizing algorithms in symmetric rings. First, we develop tight cutoffs that guarantee (1) closure in legitimate states, and (2) deadlock-freedom outside the legitimates states. We also develop a sufficient condition for convergence in silent self-stabilizing systems. Since some of our cutoffs grow with the size of local state space of processes, we also present an automated technique that significantly increases the scalability of synthesis in symmetric networks. Our technique is based on SMT-solving and incorporates a loop of synthesis and verification guided by counterexamples. We have fully implemented our technique and successfully synthesized solutions to maximal matching, three coloring, and maximal independent set problems.