The readings shared in Bluesky on 26 February 2025 are

• Big-Math: A large-scale, high-quality math dataset for reinforcement learning in language models. ~ Alon Albalak et als. #LLMs #Math
• UGMathBench: A diverse and dynamic benchmark for undergraduate-level mathematical reasoning with large language models. ~ Xin Xu et als. #LLMs #Math
• Empowering LLMs with logical reasoning: A comprehensive survey. ~ Fengxiang Cheng et als. #LLMs #Math #Reasoning