The field of large language models is moving towards enhancing their reasoning capabilities through the integration of structured knowledge representation and explicit reasoning steps. Recent research has focused on developing frameworks and benchmarks to evaluate and improve the logical reasoning abilities of large language models. Notable papers include: SLR, an automated synthesis framework for scalable logical reasoning, which enables the creation of large-scale benchmarks and achieves state-of-the-art results. The Enterprise Large Language Model Evaluation Benchmark provides a holistic evaluation framework for assessing large language model capabilities in enterprise contexts, revealing critical performance gaps and offering insights for model optimization. DynamicBench evaluates real-time report generation in large language models, demonstrating the effectiveness of its approach in storing and processing up-to-the-minute data.