The field of Large Language Models (LLMs) is moving towards improving their logical reasoning capabilities. Researchers are developing innovative tools and frameworks to evaluate and enhance these capabilities, such as systematic verifiers and synthetic data generation environments. These advancements aim to address the limitations of existing benchmarks and datasets, which often lack variable control and have narrow problem types and formats. The focus is on creating scalable and controllable tools that can provide nuanced insights into reasoning performance and enable effective reinforcement fine-tuning. Noteworthy papers include: SATQuest, which introduces a systematic verifier for evaluating and enhancing logical reasoning in LLMs. Loong, which presents an open-source framework for scalable synthetic data generation and verification across diverse reasoning-intensive domains. Saturation-Driven Dataset Generation, which leverages automated theorem proving to derive a massive, guaranteed-valid corpus of theorems for LLM mathematical reasoning.